Monthly Archives: March 2024

fig 4

Meaning of Supercritical Fluids in Pegmatite Formation and Critical-Element Redistribution

Leave a reply

DOI: 10.31038/GEMS.2024625

Abstract

This contribution shows that the water for some pegmatites comes from mantle deeps via supercritical fluids. Proofs are silicate melt inclusions with high water concentration. This concentration forms with the temperature solvus curves. In addition to such solvus curves, the distributions of a row of elements display characteristic Lorentzian curves regularly related to the solvus. Along with these characteristics, untypical, mostly spherical indicator minerals from mantle deeps are present in the granites and associated pegmatites in the upper crust regions. Nanodiamonds, graphite, coesite, and reidite belong to such minerals.

Keywords

Pegmatites, Supercritical fluids, Mantle minerals, Extreme element enrichment

Introduction

In recent years, we have seen an increasing interest in the genesis of pegmatites. The origin of this is the importance of rare elements, like Li, Be, Rb, Cs, REE, Nb, Ta, and many others, which are sometimes highly enriched in some pegmatites. That is also demonstrated by regular meetings and field trips, as well as by books [1,2] or publications on the classification of granite pegmatites [3]. Mainly in the discussion of the origin of pegmatites, those of the alkaline massifs (e.g., Lovozero Alkaline Massif/Kola peninsula; see Pekov, 2000) [4] are primarily unconsidered. In the end, the formation of miarolitic and other pegmatites is still a problem. The origin of the large Volyn chamber pegmatites (20 x 20 x 15 m [2] or the smaller miarolitic pegmatites in many granites is not entirely solved. The crucial questions are the origin of the water and the extraordinary enrichment of rare elements, which require a strong pre-enrichment. However, the origin of the water and the room-building energy is unclear, too. A typical case is the famous Sauberg tin deposit near Ehrenfriedersdorf in Saxony. Schröcke [5] has, in his notable contribution to the paragenesis of the tin deposits of the Erzgebirge, shown that the number of pegmatite veins and bodies is so large that the derivation of the necessary water cannot be the adjoining granites. The question of the origin of water for the formation of pegmatites is already unanswered. There is tacit consent that the water comes from the granite itself. Some authors ignore water as an essential component (for example, London 2008). However, water is an indispensable compound for pegmatite formation. Lindgren’s [6] model is often used to explain the formation of the different stages, starting at the liquid magmatic stage, the pegmatite stage, the pneumatolytic stage, and the hydrothermal stage at the end. However, Niggli published such a diagram (page 120) and explanation (page 122) in 1920 [7]. Fermann [8] presents his geochemical-genetic classification of granite pegmatites, which is the basis of later works in this field. In his work (1931b) [9], Fersmann suggests in Figure 5 (page 676) a direct relationship of pegmatites, pneumatolites, and hydrothermalites to the granites and has anticipated the idea of Lindgren (1937) [6]. The present authors have a contradictory opinion. At least a part of the pegmatites results from the interaction of supercritical fluids from mantle deeps with the granites at the intrusion level. Since the beginning of studies on melt inclusion in pegmatite quartz from the tin deposits, particularly on the pegmatites from the Sauberg mine in the Ehrenfriedersdorf district (Central Erzgebirge, Germany), in the years around the turn of the century, the first author often found very water-rich melt inclusions. Water content and temperature appear to be a characteristic relationship from the beginning: a typical solvus curve [10,11]. After that, such curves were also found for many other pegmatites and evolved granites (Figure 1).

fig 1

Figure 1: Pseudo-binary solvus curve for 19 different evolved granites and pegmatites worldwide. Note: each point represents the arithmetic mean of measurement on up to 100 melt inclusions [12].

Remarks to Figure 1: Values at T/TC (T in °C) = 1.0 correspond to the solvus curve’s critical point (CP). The abscissa (analytically determined water content) first approximates the melt density. Note here that the point scattering is, in a first approximation, the result of the complex interaction of volatiles (H2O, F, H3BO3), which sometimes work additively (according to Thomas and Rericha, 2023) [12]. Also, minor errors due to the inaccuracy of analytical measurements are additionally possible. For the origin of such curves, a clear answer could not given at this time. The necessary analytical technique was still in its infancy – however, the evidence of the characteristic relationship between water content and temperature increases significantly yearly. Such solvus curves were also obtained during hydrothermal diamond anvil cell (HDAC) experiments [13,14]. However, the indispensable step from experiment to nature is often missing. See to this Niggli (Figure 2a) [7].

Applying critical parameters (H2Ocrit, Tcrit, and CA-crit) displays relatively good comparability for different granite and pegmatite systems (Figure 2b). Figure 2b is a combination of Figures 1 and Figure 2a schematized.

Such Lorentzian curves are typical for many pegmatites, evolved granites, and other mineralizations, like the Habachtal emerald deposit [15-17]. Figure 2b schematically combines the solvus curves (water versus temperature) with the Lorentzian distributions of elements or anions, like SO42- (Königshain), or CO32- (Habachtal), or Rb, Cs, Be, Sn for the Ehrenfriedersdorf pegmatites and high-temperature mineralizations.

fig 2a

Figure 2a: Plots the normalized element concentration CA/CA-crit versus the normalized water concentration H2O/H2O-crit. This Lorentzian distribution is based on 216 data points (2-10 measurements of each point.

fig 2b

Figure 2b: Schematic three-dimensional diagram in the reduced coordinates (following Guggenheim, 1945 [18]; however, we do not use absolute temperatures here). X-axis: water concentration in a silicate melt (H2O) divided by the water concentration at the critical point (CP). Y-axis: temperature T (in °C) divided by the critical temperature (T-crit) at the point CP. Z-axis: element concentration CA divided by the concentration (CA-crit) at the critical point CP.

There is a systematic relationship between water concentration and temperature and a relationship between the concentration of various elements with water concentration and temperature, mainly in the form of Lorentzian-type curves. The critical point of the solvus curve coincides with the maximum of the Lorentzion distributed elements. The variation of elements showing such Lorentzian-type distribution shows instructive the high dynamic behavior of the supercritical fluids trapped in different melt inclusions. At this point, it is essential to emphasize that the supercritical fluid arriving at the solvus curve (especially at and immediately near the critical point) passes over in the critical and under-critical states (e.g., Ni et al., 2017 [19], Figure 1b in it). So, a more universal relationship is standing behind.

For about 20 elements, the first author and colleagues (e.g., Thomas et al., 2019) [15,16] could demonstrate the validity of such diagrams. However, revealing such curves was the first step to solving this puzzle standing behind the solvus curves, and the combined relationship to the enrichment of elements (the relationship between solvus and Lorentzian curves) has a deeper origin. Combining the two curve sets (pseudo-binary solvus and Lorentzian curves for each component (Figure 2), we see that the statement that the density of the silicate melt decreases is not continuous with the increase in the water content. At the critical point (CP), the concentration of some anions and cations increases rapidly; therefore, the density of the melt also shows a discontinuous, mostly sharp decrease. It is also noteworthy that the surface tension of a liquid decreases with increasing temperature and vanishes at the critical point (Guggenheim, 1945) [18]. If we look at the substantial increase of some elements at or near the critical point, the temperature increases to very high values of about 1000°C or more. From this, the primary temperatures and pressures represent supercritical conditions with typical properties: the viscosity of such fluid goes into the direction of zero, and the diffusivity increases in direction to infinity (as crucial properties). As already shown, the surface tension also vanishes at the critical point. An explanation for this behavior is only possible if we assume a supercritical fluid is coming directly from deep Earth’s mantle. We have found many proofs in crustal rocks, such as diamonds, moissanite, stishovite, coesite, etc. [20]. Traditionally, the miarolitic pegmatites in granites are directly related to their host granite [7,21-24]. Careful study of melt inclusions in many granites demonstrates that besides the typical melt inclusions, representing the granite crystallization (water concentrations between 1 and 12 % – see Johannes and Holtz, 1996) [25], there are also melt inclusions present, which depict a solvus curve with water concentrations between about 5 to more than 50% [26] and a critical point (CP) at the maximum of the solvus curve. A derivation of water directly from the crystallizing granite is complicated because pegmatite veins sometimes have root zones outside the granites.

Key Observations

An essential observation is the striking relationship between the solvus curves and the related Lorentzian element distribution around the solvus crest (Figure 2). However, these combinations are imperative but insufficient to state that the supercritical fluids come directly from deep in the mantle. The first proof came from an untypical pegmatite case: the prismatine rock from Waldheim/Saxony. This characteristic rock, with its prismatine crystals, has been discussed till now as a classic metamorphic rock [27]. The second author has intensively studied this rock from an independent standpoint. The prismatine contains high-temperature melt inclusions, which are very water-rich (see Thomas et al., 2022; Figures 9a and 9b in there) and have nothing to do with a poor metamorphic origin – they are atypical for granulite-facial minerals. Besides such H2O-rich melt inclusions, the prismatine crystals also have trapped minerals, like coesite, reidite, and others from mantle deeps. These crystal remnants are mostly spheric with a very smooth (like polished) surface, demonstrating a very fast crystallization of the prismatine. The crystallization of the prismatine host is so fast that the natural crystal habitus of the included spherical crystals is wholly suppressed. The first author has a different view of prismatine crystallization. It is a local pegmatite crystallization initiated by rising supercritical fluids from the mantle. This mantle mineral-bearing supercritical fluid brings the necessary boron from the deep. The fast-crystallizing prismatine includes the mantle minerals (smooth spheres of diamond, coesite) unchanged in the crystal matrix. Of course, the surrounding rock is of metamorphic origin (Figure 3).

fig 3

Figure 3: Pegmatitic prismatine zone in the granulite from Waldheim/Saxony

Columnar prismatine crystals (up to 3 cm long) are preferentially concentrated in zones that are more than 1 cm thick. In some rock parts, the concentration of prismatine crystals is very high [28], their Figures 1a to 1c).

In the meantime, we have found many examples with diamond, graphite, moissanite, and others in more crustal rocks (granites and pegmatite), for which we have often also seen the combination of solvus and Lorentzian curves [29-32]. For the Variscan Königshain granite near Görlitz/Lusatia, we [32] found magmatic epidote, which indicates a speedy ascent of the granite magma (700 to 1000 m/year). With the proof of diamond in zircon of the same rock samples, Thomas (2023a) [31] could show that the flow of a supercritical fluid initiates that fast ascent. Also, many quartz veins in this region can explained by fast-rising supercritical fluids. The solvus curves, the Lorentzian element distribution of some compounds (such as sulfate), and the sporadic appearance of diamonds and graphite in the high-temperature quartz and zircon support this preceding statement. It is not simple to find the high-temperature melt inclusion in the quartz by the more or less strong overprint due to the low-temperature fluid inclusions representing a multi-stage hydrothermal activity. An example is in Thomas et al. (2019) [15,16]. Another example is the steep dipping beryl-quartz veins related to the Sauberg tin deposit near Ehrenfriedersdorf. Figure 4 shows details of such a beryl-quartz vein from the Sauberg mine, and Figure 5 depicts one beryl crystal with a moissanite whisker in the centrum of that crystal.

fig 4

Figure 4: Detail of the studied quartz-beryl sample from the Sauberg mine near Ehrenfriedersdorf: Brl – green beryl, Mlb – molybdenite, Qtz – quartz [33].

fig 5

Figure 5: An about 600 μm long moissanite (SiC) whisker in the center of a beryl (Brl) crystal (crossed Nicols). Qtz – quartz [33].

This sample’s beryl and, more sporadically, quartz also contains many nanodiamonds and moissanite crystals, including and partly grown at a crustal level untypical for those crystals. This example clearly shows that unpredictable processes can work at changing the supercritical fluid into a critical/undercritical one. Nanocrystals of diamond and moissanite are germs of larger crystals. Recently, Thomas (2024a) [34] showed that in small (~2 cm) steeply inclined pegmatite veins in gneiss from drill cores near Annaberg/Erzgebirge, many nanodiamonds and graphite are present, which demonstrates that supercritical fluids rising from mantle regions are also present far away from the classic tin deposits in the Ehrenfriedersdorf region. Notably, there is a high concentration of distributed submicroscopic diamond and graphite crystallites (Thomas, 2024) in a small vertical pegmatite vein. From all observations, it follows that the composition of the supercritical fluids is highly variable from place to place. Sometimes, the typic relicts are a combination of diamond and graphite only, and sometimes, we have observed very complex parageneses (diamond, graphite, moissanite, coesite, cristobalite-X-I (high-pressure phase of cristobalite; see Černok et al., 2017) [35], and others. Careful observations of very different pegmatites and other mineralizations (Habachtal emerald deposit) have shown that nanodiamonds, graphite, and other high-pressure minerals are testimony of the supercritical fluids. Regarding the Habachtal emerald, its melt inclusions contradict the regional geology. The thesis of the involvement of supercritical fluid solves this dilemma elegantly (see Thomas et al., 2020) [17]. Also, the first author found nanodiamonds, besides graphite, in eudialyte-bearing [Na4(Ca, Ce)2(Fe2+, Mn2+)ZrSi8O22(OH, Cl)2] pegmatites in the Lovozero Massif [4] (Figure 6).

fig 6

Figure 6: Raman spectrum of graphite-diamond aggregate in nepheline in the eudialyte rock. The diamond line at 1324.8 cm-1 has a FWHM of 45.7 cm-1. The graphite main band at 1606.2 cm-1 is remarkably high.

As a rule, the diamonds are tiny, about 4 µm in diameter. Larger spherical graphite crystals are widespread (Figure 7 and Table 1).

fig 7

Figure 7: Small diamond (D) crystal with graphite (Gr) in nepheline (Ne). The crystal is 30 µm under the sample surface.

Table 1: Raman data of spherical diamond and graphite inclusions in pegmatite nepheline in eudialyte-dominant rock from Lovozero Massif (15 spherical crystals).

Phase

 Position (cm-1) FWHM (cm-1) Position (cm-1)

FWHM (cm-1)
Diamond

1333.0 ± 8.6

81.0 ± 16.2
Graphite G

1578.6 ± 9.5

80.9 ± 10.8
Graphite D2

1604.3 ± 3.0

 42.5 ± 8.0

FWHM – full width at half maximum

Interpretation

Based on innumerable results from natural samples, we could demonstrate that supercritical fluids significantly affect many mineralizations in water and energy delivery. Niggli [36] already emphasized the outstanding significance of H2O for forming pegmatites. He wrote (page 423): “Melt solutions in which water is plentiful as an essential volatile component besides large amounts of non-volatile components are called pegmatitic solutions.” However, over the origin, the reader remains in the obscure. Ni [37] gives in his introduction to advances in the study of supercritical fluids essential hints to the effects of supercritical fluids, which covers a broad spectrum between hydrous melts and aqueous solutions with an emphasis on the enrichment of large ion lithophile elements (LILE), U-Th-Sr and REE. Sun et al. [38] show that supercritical fluids’ speciation and transport properties are essential for understanding their behavior in the Earth’s interior. According to our studies, supercritical fluids connect the deep mantle and the crust by delivery of energy, water, and various elements, e.g., coming from subducted zones. Our studies answer those open questions regarding the origin of water. There is already a lot to solve all problems related to the origin of water and the processes connected with the transition from the supercritical to the under-critical stage of water. The enrichment processes in this state, indicated by the Lorentzian type of element distribution, are a further open problem. Element enrichment in a relatively small room by the factors of 104 to 105 is not seldom. Is there communication between the particles, like the quantum entanglement in the supercritical state?. We have seen that mantle minerals like nanodiamonds are already present in the old Volyn pegmatites (1760 Ma) [39], the pegmatites from the Lovozero Massif of Devonian age and the younger ones like the Variscan Erzgebirge and the Königshain Massif [20], Furthermore, it follows the question of the time and intensity of the ascending supercritical fluid: sporadic or continuous?

Acknowledgment

We thank many coworkers and other scientists in the past who have contributed to the supercritical fluids and their proof in natural samples. A special thanks go to Huaiwei Ni (Hefei, China), Yicheng Sun (Nanjing), and Jim Webster (New York).

References

  1. London D (2008) Pegmatites. The Canadian Mineralogist, Special Publication 10.
  2. Pavlishin VI, Dovgyi SA (2007) Mineralogy of the Volynian Chamber Pegmatites, Ukraine. Volodarsk- Mineralogical Almanac 12.
  3. Müller A, Simmons W, Beurlen H, Thomas R, Ihlen PM, et al. (2018) A proposed new mineralogical classification system for granitic pegmatites – Part I: History and the need for a new classification. The Canadian Mineralogist 56: 1-25.
  4. Pekov IV (2000) Lovozero Massif: History, Pegmatites, Minerals. Moscow.
  5. Schröcke H (1954) Zur Paragenese erzgebirgischer Zinnerzlagerstätten. Neues Jb. Abh 87: 33-109.
  6. Lindgren W (1937) Succession of minerals and temperatures of formation in ore deposits of magmatic affiliation. American Institute of Mining, Metallurgical, and Petroleum Engineers. Technical Publication 126.
  7. Niggli P (1920) Die Leichtflüchtigen Bestandteile im Magma. B.G. Teubner Verlag Leipzig.
  8. Fersmann A (1931a) Über die geochemische-genetische Klassifikation der Granitpegmatite. Zeitschrift für Kristallographie, Mineralogie und Petrographie, Abteilung B Mineralogische und Petrographische Mitteilungen 41: 64-83.
  9. Fersmann A (1931b) Geochemische Diagramme. Neues Jb. Mineral. Etc. Abh.
  10. Thomas R, Webster JD, Heinrich W (2000) Melt inclusions in pegmatite quartz: complete miscibility between silicate melts and hydrous fluids at low pressure. Mineral, Petrol. 139: 394-401.
  11. Thomas R, Förster HJ, Heinrich W (2003) The behaviour of boron in a peraluminous granite-pegmatite system and associated hydrothermal solutions: a melt and fluid-inclusion study. Mineral, Petrol. 144: 457-472.
  12. Thomas R, Rericha A (2023) The function of supercritical fluids for the solvus formation and enrichment of critical elements. Geology, Earth and Marine Sciences (GEMS) 5: 1-4.
  13. Shen AH, Keppler H (1997) Direct observation of complete miscibility in the albite-H2O system. Nature 385: 710-712.
  14. Bureau H, Keppler H (1999) Complete miscibility between silicate melts and hydrous fluids in the upper mantle: experimental evidence and geochemical implications. Earth and Planetary Letters 165: 187-196.
  15. Thomas R, Davidson P, Appel K (2019) The enhanced element enrichment in the supercritical states of granite-pegmatite systems. Acta Geochim 38: 335-349.
  16. Thomas R, Davidson P, Rericha A, Tietz O (2019) Eine außergewöhnliche Einschlussparagenese im Quarz von Steinigwolmsdorf/Oberlausitz. Berichte der Naturforschenden Gesellschaft der Oberlausitz 27: 161-172.
  17. Thomas R, Davidson P, Rericha A (2020) Emerald from the Habachtal: new observations. Mineralogy and Petrology 114: 161-173.
  18. Guggenheim EA (1945) The principle of corresponding states. The Journal of Chemical Physics 13: 253-261.
  19. Ni H, Zang L, Xiong X, Mao Z, Wang J (2017) Supercritical fluids at subduction zones: Evidence, formation condition, and physicochemical properties. Earth-Science Reviews 167: 62-71.
  20. Thomas R (2023) The Königshain Granite: Diamond inclusion in zircon. Earth Mar. Sci. 5: 1-4.
  21. Niggli P (1937) Das Magma und seine Produkte. Akademische Verlagsgesellschaft M.B.H. Leipzig.
  22. Fersman AE (1934) Geochemistry Vol. II, Leningrad, 1-354 (in Russian).
  23. Fersman AE (1937), Geochemistry Vol. III, Leningrad, 1-503 (in Russian).
  24. Fersman AE (1961) Les Pegmatites. Librairie Universitaire Uystpruyst, Louvain. Tome I, 163 p.; Tome II, 164-437 p.; Tome III, 438-739.
  25. Johannes W, Holtz F (1996) Petrogenesis and Experimental Petrology of Granitic rocks. Springer Verlag.
  26. Thomas R, Förster HJ, Rickers K, Webster JD (2005) Formation of extremely F-rich hydrous melt fractions and hydrothermal fluids during differentiation of highly evolved tin-granite magmas: a melt/fluid-inclusion study. Mineral. Petrol. 148: 582-601.
  27. Grew ES, Thomas R (2021) Coesite and relict diamond in prismatine and pyrope from Waldheim, Germany: New evidence for deep burial of the Saxony Granulitgebirge and its implications for the Bohemian Massif of central Europa. Research Seminar at the University of Maine-Orono, September 15.
  28. Thomas R, Davidson P, Rericha A, Recknagel U (2022c) Water-rich coesite in prismatine-granulite from Waldheim/Saxony. Veröffentlichungen Museum für Naturkunde Chemnitz. 45: 67-80.
  29. Thomas R, Davidson P, Rericha A, Recknagel U (2022a) Discovery of stishovite in the prismatine-bearing granulite from Waldheim, Germany: A possible role of supercritical fluids of ultra-high-pressure origin. Geosciences 12: 196, 1-13.
  30. Thomas R, Davidson P, Rericha A, Recknagel U (2022b) Supercritical fluids conserved as fluid and melt inclusions in quartz from the Sheba-Gold Mine, Barberton, South Africa. Aspects in Mining & Mineral Science 10: 1193-1196.
  31. Thomas R, Davidson P, Rericha A, Recknagel U (2023a) Ultrahigh-pressure mineral inclusions in a crustal granite: Evidence for a novel transcrustal transport mechanism. Geosciences 94: 1-13.
  32. Thomas R, Davidson P (2016) Origin of miarolitic pegmatites in the Königshain granite/Lusatia. Lithos 260: 225-241.
  33. Thomas R, Recknagel U, Rericha A (2023b) A moissanite-diamond-graphite paragenesis in a small beryl-quartz vein related to the Ehrenfriedersdorf deposit, Germany. Aspects in Mining & Mineral Sience 11: 1310-1319.
  34. Thomas R (2024a) Graphite and diamond-rich pegmatite as a small vein in a gneiss drill core from the Annaberg region Erzgebirge, Germany. Geology, Earth and Marine Sciences (GEMS) 6: 1-4.
  35. Černok A, Marquardt K, Caracas R, Bykova E, Habler G, et al. (2017) Compressional pathways of -cristobalite, structure of cristobalite X-I, and towards the understanding of seifertite formation. Nature Communications 8: 1-12.
  36. Niggli P (1954) Rock and Mineral Deposits. W.H. Freeman and Company, San Francisco.
  37. Ni H (2023) Introduction to advances in the study of supercritical geofluids. Science China Earth Sciences 66: 2391-2394.
  38. Sun Y, Liu X, Lu X (2023) Structures and transport properties of supercritical SiO2-H2O and NaAlSi3O8-H2O-fluids. American Mineralogist 108: 1871-1880.
  39. Thomas R (2024b) Another fluid inclusion type in pegmatite quartz: Complex organic compounds. Geology, Earth and Marine Sciences (GEMS) 6: 1-5.
FIG 1

Mind Genomics and Today’s Realpolitik: Considering the ‘Invasion’ of Single Young Men at the US Southern Border from the Point of View of What Mind-sets Might Exist and What to Consider

Leave a reply

DOI: 10.31038/JIAD.2024114

Abstract

A combination of Mind Genomics to understand motivation coupled with Idea Coach (artificial intelligence module within Mind Genomics) was used to create synthetic mind-sets which might describe young males illegally crossing the US southern border. The paper shows how AI can provide information to spur critical thinking when provided with a description of the situation and the motivation for the illegal crossing. The authors suggest that the world of law enforcement might benefit by using these procedures to facilitate critical thinking.

Introduction

As of this writing, February 2024, the United States is experiencing a never-before-situation at its borders. The southern borders of the United States, especially those in Texas, are being inundated by migrants, many of whom are young, unaccompanied men, who slip into the United and end up disappearing inside the United States. Many of these people disappear entirely. Others go to court and are allowed to say pending their case.

Up to now, the material just presented recognizes a problem emerging, namely the escape of many unregistered aliens through what has turned out to be an exceptionally porous border, manned by seriously understaffed border patrols and immigration officers. The consequence is not unexpected. Many people believe that through such unmonitored immigration, there is a good chance an at alien army my be coming in, this army not necessary friends o the United States, the company into which they are disappearing.

The objective of this paper is to demonstrate how AI can be used to formulate hypotheses about the nature of what might be happening at the southern border, and then to demonstrate the change in AI-based ‘conclusions’ when the motive for the invasion includes intending to harm the United States.

History of the Approach

The tool used here is Idea Coach, an AI-empowered program embedded in the Mind Genomics platform [1]. In turn, Mind Genomics is a platform which specializes in the analysis of human judgments, doing so by presenting the respondent (survey taker) with a variety of vignettes, and for each vignette, obtaining a rating on defined scale. The vignettes themselves are combinations of simple, easy-to-read statements, called elements. The vignettes are constructed according to a plan known as an experimental design. Each respondent evaluates a unique set of 24 vignettes [2]. The ratings are then transformed to a simple Yes/No scale. The final steps are to use simple OLS (ordinary least-squares) regression at the level of each respondent to relate the presence/absence of the elements to the transformed ratings. The coefficients of the equations become the tool to understand the mind of the respondent. This simple analysis shows immediately what elements ‘drive’ the ratings, and by so doing reveal the underlying mind of the respondent with regard to the specific topic. The final analysis clusters the respondents into different groups based upon similar patterns of coefficients.

The foregoing approach has been embedded into the aforementioned Mind Genomics platform (www.bimileap.com). The original format of Mind Genomics required that the user provide a set of four questions ‘which tell a story’, and then ‘four answers to each question.’ It was the answers which the respondent evaluated, after these answers (now called elements) were mixed and matched to create the vignette.

AI was necessary to help the user think of questions and answers. Over the period of several years, it became increasingly obvious that users of Mind Genomics both liked the approach but were terrified of the requirement to come up with questions and answers. Figure 1 shows the request to provide four questions. This request was a wall to many prospective users because, quite simply, it was daunting. People were often good at answering questions but not at formulating questions to tell a story.

FIG 1

Figure 1: The set-up screen for Mind Genomics studies. Panel A requires the user to provide four questions which tell a story. Panel B shows the Idea Coach, giving the user the opportunity to describe the topic, and turn receive 15 questions generated by AI.

The incorporation of AI into the process of questions and answers increased the acceptance of the Mind Genomics platform, for at least four reasons:

  1. The process no longer stymied the user. A simple ‘squib’ in Figure 1, Panel B, sufficed to generate 15 questions.
  2. The same process occurred in the generation of elements. Once the user selected four questions and put those questions into the template (Figure 1, Panel A), the Idea Coach was able to return 15 answers to each question selected.
  3. The process was rapid, with the suggested sets of 15 questions or 15 answers to a question returning in about 20 seconds.
  4. The user could edit the squib to change the nature of the questions, or edit the selected questions to change the nature of the answers

The incorporation of AI as Idea Coach ended up producing Idea Books, compilations of questions produced in response to the squib (Figure 1, Panel B), as well as compilations of answers produced in response to each question. Each page in the Idea Book corresponded to one iteration, whether the 15 suggested questions resulting from a request written out in the squib, or 15 answers resulting from the selection of a question. It was not unusual to generate Idea Books of 10+ tabs.

In addition to the sets of questions or answers on each page, the AI was given the task of summarizing the material on each page, viz., the questions or answers. The result was other insights, such as key ideas, themes, perspective, what is missing, interested audiences, opposing audiences, and innovations. Each of the foregoing was given its own section on the page in bold type, and then the relevant AI summarization provided.

The foregoing process required about 30 minutes in total from beginning the set up of the Mind Genomics experiment to the creation of a book with say 20 pages. The process itself was quick, the results were easy to obtain, and the iterations themselves became a source of learning, the Idea Book turning into a resource book for further work.

Over time, and as the process became easier, the process first became rigid as practitioners using Mind Genomics followed the path laid out, with simple questions posed to the AI embedded in Idea Coach. The only modifications during the early days of Idea Coach, the year 2022-2023, was the expansions of the question, so that the questions would have a certain number of word (~ 10-15), that the questions would be interesting, and that the questions could be understood by a young person. The same ‘editing’ of requests was done for the questions themselves in order to generate answers which ‘were not lists, but rather statements which could lead to a discussion’. All of these were style questions, rather than substance questions. The happy consequence was that the Idea Book was richer in content, the questions and answers more instructive, and the process enjoyable to the user, who could practice writing different requests about the style and structure of the output to be generated by Idea Coach

Advancing Insights through Deeper Interactions with AI through the Idea Coach

During the course of working with Idea Coach, author Mulvey expanded the nature of the squib, and generated unexpected and deeper results. Rather than simply specifying the nature of the question or answer in terms of style (viz., number of words, age of reader, style to engage the reader rather than list options), Mulvey added a request into the squib. That request was to provide some additional answers to the question. That is, the squib or input to the AI embedded in Idea Coach contained a request for additional structure in the question, rather than just a question alone. The approach is conceptually similar to the creation of synthetic data, in this case synthetic mind-sets [3].

Idea Coach returned with unexpectedly deeper questions. The output to this ‘expanded request’ was more like a summary of a situation than simply suggested questions. Exploration of alternative ways to expanded the input to Idea Coach quickly revealed that the AI could be requested to a far deeper analysis. And so, the approach was born which lies at the foundation of this paper.

The paper itself explores how the Idea Coach provides more information when given a detailed instruction. Table 1 shows the text as it appears in the Idea Book returned to the respondent. The top section in bold shows what the user types into the squib. The rest of the table shows what it returned to the user by Idea Coach when the Idea Book is completed.

Table 1: Output from Idea Coach for the first request, where the hostile goals of the border crossing men are not revealed.

TAB 1(1)

TAB 1(2)

TAB 1(3)

The first part of the input to Idea Coach is simple and direct, describing what is happening., and a simple request for suggestions: The topic is: Invasion of the United States at the southern border cross by unmarried young men of military age. How can we prevent these people from starting massacres at unarmed gatherings throughout the United States if they are truly ‘invading us’.

The second part of the input to Idea Coach is the set-up hypothesis that there exist three mind-sets. These mind-sets are not named, and indeed no information is given about any conjectures regarding the mind-sets. It will be the job of the AI in Idea Coach to suggest the mind-sets. We will see that the suggested mind-sets returned by AI ends up concurring with additional information provided by the user. The actual text is straightforward: We believe that there are three different mind-sets of these young men. The user can change the number of mind-sets and add more information about the mind-sets. Each change generates a new set of responses, further serving as an educational and preparedness tool for the user. and can be modified by the user to see what happens when the number of mind-sets is changed.

The third part of the input to Idea Coach is the instruction to answer a set of eight specific questions for each to-be-named mind-set:

For each mind-set in turn, answer these specific questions:

  1. What is the name of the mind-set
  2. What is the goal of the mind-set specifically with regard to being in the United States
  3. What are they likely to say to an immigration official when they are caught
  4. What will convince them to go through legal channels to become regular citizens
  5. How can we recognize them… give four indications to help recognize the mind set to which they belong
  6. If unrecognized, what will they likely do in three months after they have entered illegally into the United States
  7. If unrecognized, what will they likely do in six months after they have entered illegally into the United States
  8. What will make them want to identify themselves to officials in the United States

Do the above answers for each mind-set separately, answering all questions 1-8 for each mind-set in turn

The fourth and final instruction to Idea Coach focuses on the style of the suggestion to be given by Idea Coach:

Make the answers interesting to read, and easy to talk about to other people

Make the answers as realistic as possible

Recognize that the answers will be shared with officials in the United States

Section B in Table 1 shows the immediate work-product of the AI embedded in Idea Coach. Section B returns with 30 seconds. Generally, but not always, Idea Coach provides precise sets of answers, following the instructions written in the squib. The user is free to re-run the Idea Coach many times, changing the squib desired. (That change will be shown in Table 2, where the input to Idea Coach will add information about the migrants ‘wanting to harm America’.) Each iteration of Idea Coach generates a mix of new ideas and old ideas

Section C in Table 1 shows a set of summaries created by AI, with these summaries taking into account all of the information in Sections A and B, respectively. The summarizing queries in Idea Coach are fixed, but the information in each iteration tends to be partially unique, so that the summarizations will differ from iteration to iteration.

When looking at different sections of Table 1 the reader should keep in mind that within less than a minute the user has gone from a set of questions to a set of answers, and in a few more minutes from a set of answers to a set of summarized generalities. It is also important to keep in mind that the typical Idea Book does not stop with one or two iterations, but may go on for dozens of iterations, each done without much effort, each done to satisfy one’s curiosity about a particular issue, not only about the general topic.

Iterating and Adding Information about Motives for a More Targeted Analysis

As stated above, a key benefit of the Mind Genomics approach as empowered by the Idea Coach emerges from the ability to modify the squib or request given to the underlying AI. Table 1 showed the three mind-sets without any specification of who the mind-sets are, other than the general concern about a possible massacre in a public place.

Tables 2-5 show the descriptions of the mind-sets developed by AI when the invading males are further specified as to their assumed motives. The issue now is to see how these three mind-sets are described. For the purposes of this paper, the focus is simply on the types of descriptions which emerge when additional information is provided to the AI embedded in Idea Coach. In a sense the descriptions of the mind-sets generated in Tables 1-5 can be looked at an exploration of how AI can put features onto basic descriptions.

Table 2: Three mind-sets emerging from AI when the motivation is to start a family in the United States

TAB 2

Table 3: Three mind-sets emerging from AI when the motivation is to start working and then become citizens

TAB 3(1)

TAB 3(2)

Table 4: Two mind-sets emerging from AI when the motivation is to get jobs and bring in their family living in poverty

TAB 4

Table 5: Three mind-sets emerging from AI when the motivation is to harm the United States with gang warfare

TAB 5

Discussion and Conclusions

The ingoing rationale for this study was to demonstrate that a new opportunity to understand behavior has evolved from the incorporation of AI through Idea Coach into the basic thrust of Mind Genomics. The original objectives of Mind Genomics focused on understanding motivations and decision ‘rules’ for different types of people, with these rules emerging from material taken from the granular, everyday world. The development of AI made it possible to accelerate the process by producing a way to help people ask questions and create answers for that question.

The introduction of AI also made it possible to go into directions not thought of before, or if thought of, then consigned to the world of the theoretical. We are talking here about ‘what if’ questions. What if we could ascribe basic motivations to people, almost making the task which involves synthetic people, rather than real people. This paper shows what can be done by creating synthetic people, simply by telling AI that there are three mind-sets, giving some background, and then systematically varying some aspect of that background. Tables 1-5 show what happens when the user moves from no motivation stated to a variety of different motivations.

Up to now a major focus of AI in law enforcement has been to detect patterns in the transfer of money and other objects, almost the combination of Big Data and the Internet of Things [4-7]. The use of AI to construct synthetic mind-sets for law may be in progress but is not yet mainstream. On the other hand, the use of AI to construct synthetic people for surveys is beginning to become mainstream, at least in the market research community. There is every reason to assume that the use of AI to construct scenarios and synthetic people, as well as synthetic mind-sets, will become mainstream, and perhaps even significant. To the degree that the approach presented here becomes a tool for critical thinking in law and public policy, we may expect to see the approach presented here proliferate and improve thinking as well as public policy.

References

  1. Moskowitz H, Todri A, Papajorgji P, et al. (2023) Sourcing and vetting ideas for sustainability in the retail supply chain: The Contribution of artificial intelligence coupled with Mind Genomics. International Journal of Food System Dynamics 14: 367-380.
  2. Gofman A & Moskowitz H (2010) Isomorphic permuted experimental designs and their application in conjoint analysis. Journal of Sensory Studies 25: 127-145.
  3. Raghunathan TE (2021) Synthetic Data, Annual Review of Statistics and Its Application 8, 129-140.
  4. Borum R (2020) Scientific and technological advances in law enforcement intelligence analysis. Science Informed Policing 99-121.
  5. Rademacher T (2020) Artificial intelligence and law enforcement. In: Regulating Artificial Intelligence, (ed, T. Wischmeyer & T, Rademacher), Springer Link 225-254.
  6. Watney M (2019) Law Enforcement Use of Artificial Intelligence for Domestic Security: Challenges and Risks. In Proceedings of the European Conference of Artificial Intelligence and Robotics (ECIAR), Oxford, UK, November 341-348.
  7. Yavorsky MA and Mikheeva SV (2022) The Use of Artificial Intelligence Technologies and Big Data in Law Enforcement. In Proceedings of the International Conference Engineering Innovations and Sustainable Development 669-675. Springer International Publishing.

Gaza as a Middle East Singapore – Enhanced Visioning of Opportunities Suggested by AI (Socrates as a Service)

Leave a reply

DOI: 10.31038/JIAD.2024113

Abstract

The project objective was to explore the opportunity of Gaza as a new Singapore, using AI as th source of s suggestions to spark discussion and specify next steps. The results shows the ease and directness of using AI as a ‘expert’ when the materials presented here were developed through the Mind Genomics platform BimiLeap.com and SCAS (Socrates as a Service). The results show the types of ideas developed, slogans to summarize those ideas, AI-developed scales to profile the ideas on a variety of dimensions, and then five ideas expanded in depth. The paper finishes with an analysis of types of positive versus negative responses to the specific solutions recommended, allowing the user to prepare for next steps, both to secure support from interested parties, and to defend against ‘nay-sayers’.

Introduction

Europe had been devastated by Germany’s World War II invasion. The end of World War II left Germany in ruins. The Allies had deported many Germans, leveled Dresden, destroyed the German economy, and almost destroyed the nation. From such wreckage emerged a more dynamic Germany with democratic values and a well-balanced society. World War II ended 80 years ago, yet the effects are being felt today. The Israel-Hamas battle raises the issue of whether Gaza can be recreated like Germany, or perhaps more appropriately Singapore Can Gaza become Middle East Singapore? This article explores what types of suggestions might work for Gaza to become another Singapore, with Singapore used here as a metaphor. Businesses have already used AI to solve problem [1,2]. The idea of using AI as an advisor to help with the ‘social good’ is alo not new [3-5]. What is new is the wide availability of AI tools, such as Chat GPT and other larger language models [6]. ChatGPT are instructed to provide replies to issues of social importance, specifically how to re-create Gaza as another Singapore. As we shall see, the nature of a problem’s interaction with AI also indicates its solvability, the acceptance by various parties and the estimated years to completion

The Process

First, we prompt SCAS, our AI system, to answer a basic question. The prompt appears in Table 1. The SCAS program is instructed to ask a question, respond in depth, surprise the audience, and end with a slogan. After that, SCAS was instructed to evaluated the suggestion rate its response on nine dimensions, eight of which were zero to 100, and the ninth was the length of years it would take this program or idea to succeed.

Table 1: The request given to the AI (SCAS, Socrates as a Service)

tab 1

The Mind Genomics platform on which this project was run allows the user to type in the request and press the simple button and within 15 seconds the set of answers appears. Although the request is made for 10 different solutions, usually the program returns with far fewer. To get more ideas, we simply ran the program for several iterations, meaning that we just simply pressed the button and ran the study once again. Each time the button was pressed by the user, the program returned anew with ideas, with slogans and with evaluations. Table 2 shows 37 suggestions. emerging from nine iterations, the total time taking about 10 minutes. The importance of many iterations is the AI program does all the ‘thinking,’ all the ‘heavy lifting’. Table 2 is sorted by the nature of the solution, the categories for the sorting provided by the user. Each iteration ends up generating a variety of different types of suggestions, viz., some appropriate for ‘ecology’, others for ‘energy’, other for ‘governance’ and so forth. Each iteration came up with an assortment of different suggestions. Furthermore across the many iterations, even the ‘look’ of the output change. Some outputs comprised detailed paragraphs, other outputs comprised short paragraphs. Looking at the physical format gave a sense that the machine seemed to be operated by a person whose attention to details was oscillating from iteration to iteration. Nonetheless, the AI did generate meaningful questions, seemingly meaningful answers to those questions, and assigned the ratings in a way that a person might. The table is sorted by the types of suggestions. Thus the same topic (e.g., ecology) may come up in different iterations, and with different content. The AI program does not make any differentiation, but rather seems to behave in a way that we would call ‘does whatever comes into its mind.. It can be seen that by doing this 10, 20, 30 times and having two or three or four suggestions for each, the user can create a large number of alternative solutions for consideration. Some of these will be, of course, duplicate. Many will not. A number will be different responses, different points of view, different things to do about the same problem, such as economy.

Table 2: The suggestions emerging from eight iterations of SCAS, put into a format which allows these suggestions to be compared to each other.

tab 2(1)

tab 2(2)

tab 2(3)

tab 2(4)

tab 2(5)

Expanding the Ideas

The next was to selected five of the 37 ideas, and for each of these five ideas instructed AI (SCAS) to ‘flesh out’ the idea with exactly eight steps. Table 3 shows the instructions. Tables 4-7 show the four ideas, each with the eight steps (left to SCAS to determine). At the bottom of each table are ‘ideas for innovation’ provided by SCAS when it summarizes the data at the end of the study. The Mind Genomics platform ‘summarizes’ each iteration. One of the summaries is the ideas for innovation. These appear at the bottom section of each of the five sections, after the presentation of the eight steps.

Table 3: The prompts to SCAS to provide deeper information about the suggestion previously offered in an iteration.

tab 3

Table 4: The eight steps to help Gaza become a tourism hotspot like Singapore

tab 4

Table 5: The eight steps to make Gaza ensure inclusive and sustainable growth for all its residents

tab 5

Table 6: The eight steps to help Gaza improve its economy and living standards

tab 6

Table 7: The eight steps to help Gaza promote entrepreneurship and small business development

tab 7

The Nature of the Audiences

Our final analysis is done by the AI, the SCAS program in the background after the studies have been complete. The final analysis gives us a sense of who the interested audiences might be for these suggestions and where we might find opposition. Again, this material is provided by the AI itself and the human prodding. Table 8 shows these two types of audiences, those interested versus those opposed, respectively.

Table 8: Comparing interested vs opposing audiences

tab 8

Relations among the ‘Ratings’

The 37 different suggestions for many topic areas provides us with an interesting set of ratings assigned by the AI. One question which emerges is whether these ratings actually mean anything. Some ratings are high, some ratings are low. There appears to be differentiation across the different rating scales and within a rating scale across the different suggestions developed by AI. That is, across the 37 different suggestions, each rating scale shows variability. The question is whether that is random variability, which makes no sense, or whether that is meaningful variation, the meaning of which we may not necessarily know. It’s important to keep in mind that each set of ratings was generated entirely independently, so there is no way for the AI to try to be consistent. Rather, the AI is simply instructed to assign a rating. Table 8 shows the statistics for the nine scales. The top part shows the three decriptive statistics, range, arithmetic mean, and standard error of the meant. Keeping in mind that these ratings were assigned totally independently for each of the 37 proposed solutions (Table 2), it is surprising that there is such variability. The second analysis shows something even more remarkable. A principal components factor analysis [7] enables the reduction of the nine scales to a limited set of statistically independent ‘factors’. Each original scale correlates to some degree with these newly created independent factors. Two factors emerged. The loadings of the original nine scales suggest that Factor 1 are the eight scale of performance as well as novelty, whereas Factor 2 is years to complete. The clear structure generated across independent ratings by AI of 37 suggestions is simply very clear, totally unexpected, and at some level quite remarkable. At the very least, one might say that there is hard-to-explain consistency (Table 9).

Table 9: Three statistics for the nine scales across 37 suggested solutions

tab 9

Discussion and Conclusions

This project was undertaken in a period of a few days. The objective was to determine whether AI could provide meaningful suggestions for compromises and for next steps, essentially to build a Singapore from Gaza. Whether in fact these suggestions will actually find use or not is not the purpose. Rather the challenge is to see whether artificial intelligence can become a partner in solving social problems where the mind of the person is what is important. We see from the program that we used, BimiLeap.com and its embedded AI, SCAS, Socrates as a Service, that AI can easily create suggestions, and follow up these suggestions with suggested steps, as well as ideas for innovation. These suggestions might well have come from an expert with knowledge of the situation, but in what time period, at what cost, and with what flexibility? All too often we find that the ideation process is long and tortuous Our objective was not to repeat what an expert would do, but rather see whether we could frame a problem, Gaza as a new Singapore, and create a variety of suggestions to spark discussion and next steps, all in a matter of a few hour. The potential of using artificial intelligence to help spark ideas is only in its infancy. There’s a good likelihood that over the years as AI becomes quote smarter and the language models become better, suggestions provided by AI will be far more novel, far more interesting. Some of the suggestions are interesting, although many suggestions are variations on the ‘pedestrian’. That reality not discouraging but rather encouraging because we have only just begun. There’s a clear and obvious result here that with the right questioning, AI can become a colleague spurring on creative thoughts. In the vision of the late Anthony Gervin Oettinger of Harvard University, propounded 60 years ago, we have the beginnings of what he called T-A-C-T, Technical Aids to Creative Thought [8]. Oettinger was talking about the early machines like the EDSAC and programming the EDSAC to go shopping [9,10]. We can only imagine what happens when the capability shown in this paper falls into the hands of young students around the world who can then become experts in an area in a matter of days or so. Perhaps solving problems, creative thinking, and even creating a better world will become fun rather than just a tantalizing dream from which one reluctantly must awake.

References

  1. Kleinberg J, Ludwig J, Mullainathan S (2016) A guide to solving social problems with machine learning. In: Harvard Business Review.
  2. Marr B (2019) Artificial Intelligence in Practice: How 50 Successful Companies Used AI and Machine Learning To Solve Problems. John Wiley & Sons.
  3. Aradhana R, Rajagopal A, Nirmala V, Jebaduri IJ (2024) Innovating AI for humanitarian action in emergencies. Submission to: AAAI 2024 Workshop ASEA.
  4. Floridi L, Cowls J, King TC, Taddeo M (2021) How to design AI for social good: Seven essential factors. Ethics, Governance, and Policies in Artificial Intelligence. Pg: 125-151. [crossref]
  5. Kim Y, Cha J (2019) Artificial Intelligence Technology and Social Problem Solving. In: Koch F, Yoshikawa A, Wang S, Terano T (eds) Evolutionary Computing and Artificial Intelligence. GEAR 2018. Communications in Computer and Information Science, vol 999. Springer, Singapore.
  6. Kalyan KS (2023) A survey of GPT-3 family large language models including ChatGPT and GPT-4. Natural Language Processing Journal.
  7. H. 7 Williams LJ (2010) Principal component analysis. Wiley Interdisciplinary Reviews: Computational Statistics 2: 433-459.
  8. Bossert WH, Oettinger AG (1973) The Integration of Course Content, Technology and Institutional Setting. A Three-Year Report, 31 May 1973. Project TACT, Technological Aids to Creative Thought.
  9. Cordeschi R (2007) AI turns fifty: revisiting its origins. Applied Artificial Intelligence 21: 259-279.
  10. Oettinger AG (1952) CXXIV. Programming a digital computer to learn. The London, Edinburgh, and Dublin Philosophical Magazine and Journal of Science 43: 1243-1263.
FIG 1

Mind Genomics and Today’s Realpolitik: Considering the Conflict between Texas and the Federal Government from the Point of View of What Mindsets Might Exist and What to Consider

Leave a reply

DOI: 10.31038/JIAD.2024112

Abstract

This paper deals with how AI can summarize issues emerging from the newly developing confrontation between Texas and the US Federal Government regarding unrestricted migration across the southern US border. The paper shows how AI embedded in the Mind Genomics platform can provide a mechanism whereby a person can profoundly explore subjective thinking about a topic. The paper presents the verbatim result of an easy-to-understand, easy-to-do, and very affordable investigation, one which required less than 36 to conceptualize, implement, understand, and ‘work up’ for publication. The combination of Mind Genomics and AI in the user-friendly Idea Coach promises a new capability to democratize critical thinking, world-wide.

Introduction

Since the informal opening up of the southern US Border, the news sources have been filled with innumerable accounts of migrants coming across the border and escaping into the United States. The issues involved in this wholesale unwanted, unplanned for and frankly dangerous entrance of hordes of strangers into the US has erupted into conflicts between states and now a conflict between the state of Texas and the US Federal Government. The issue is so grave that the government in Texas is announcing that it will maintain its protective system of wires, in defiance of the United States Federal Government. As one might expect, the issues are so important, so severe that within the past month alone (January 2024) dozens of treatments of the issue have appeared in Google Scholar, remarkable for a topic [1-6]. The sheer volume of published material on the topic, the continuing ‘buzz’ of the social media, the incessant reporting by today’s media giants, and the unspoken concerns of the ordinary citizen suggest the need for a better way to understand the topic. It is to address that need that this paper was written, more from the desire to see how far AI could go to help the citizen ‘understand’, as well as how far AI could go to make a scholar contribution to civil society. This ‘experiment in AI’ looks at the situation from the point of view of what might be called the ‘Mind of Texas’. The underlying structure of the AI study is to present the situation, then suggest that there are different mind-sets about this topic, mind-sets from the point of Texas, but not to name these mind-sets. Rather, the AI is instructed to name the mind-set, state what the mind-set believes to be the proper action, how does the mind-set support that belief, and then assign a rating to the strength of the case that Texas is making with that mind-set.The important thing to keep in mind is that the effort represents a new way of thinking about AI and political issues. Rather than summarizing information to answer a specific question, AI is tasked with identifying new to the world groups, not given anything. Furthermore, for each group, the AI is tasked with finding out how they think, and finally and most uniquely, estimating the strength of the argument.

The Contribution of Mind Genomics and the Impetus of AI to Solve the Problem of Critical Thinking

The origin of the material reported here can be traced to the emerging science of Mind Genomics, colloquially the ‘science of the everyday’. The inspiration for Mind Genomics was the desire to understand how ordinary people thought about the ordinary events of their day, the granular aspects of their quotidian existence to put in a way which may have more gravitas. Cognitive psychology recognizes that people make decisions in different ways but focuses on the person making the decisions rather than on the specific topics about which the decisions are made. Mind Genomics works with combinations of elements, these elements being simple descriptions of aspects of a thing or a situation. The combines these elements into small vignettes, combinations of elements, doing so according to a scheme which prescribes the precise combinations. Respondents, or in other disciplines so-called ‘survey-takers’ then read each vignette and rate the vignette, that combination of messages. The respondent may read several dozen vignettes, one vignette at a time, and then rate the vignette on a scale. Once the data are collected, the researcher deconstructs the rating assigned to the vignette into the part-worth contribution of each of the elements. To the respondent the task is simple, although often boring. The elements are combined in different ways for the respondents, so that each respondent evaluates a different set of vignettes. The mathematical structure of the vignettes is the same form one respondent to the other. The only thing which changes is the particular set of combinations. This ability to give people different combinations of the same elements, but in reality the underlying structure is the same, is a feature of Mind Genomics. When doing the task the respondent, having been presented with these seemingly random combinations, stops trying to figure out the ‘underlying game’, stops trying to outguess the researcher, and ends up simply guessing, which turns out to be what was desired in the first place. The happy result is that the way the respondent thinks about the topic ends up being revealed by statistics (OLS or ordinary least-squares regression). The system works because it mimics nature. Nature continues to present to us vignettes, combinations of simple stimuli. In order to survive we have to pick out through this blooming, buzzing confusion in the words of the eminent 19th century Harvard psychologist, William James. The desire to understand how different elements drive the response is attractive to researchers because these elements possess ‘meaning.’ When the researcher presents combinations of phrases to the respondent, the combination does signify something, and elicits a responses, even though the combinations may be put together in what seems to be a random way. When the researcher combines the phrases according to an underlying ‘experimental design’ which specifies the particular combinations, there is nothing in the design which forces a coherent meaning. Yet, for the most part, respondents have no problem finding meaning in the combination. When asked to rate the combination, the vignette, most respondents may be at first perplexed about what to do, but within one or two exposures the respondents cease being perplexed and go about the task of reading and rating the vignettes. The behavior is almost like ‘grazing’ for meaning. That is, the respondent looks at the vignette, extracts the information needed to decide, assigns a rating and moves on. From literally thousands of studies following the use of prescribed experimental designs and inexperienced panelists who don’t know what to do, the data which emerge makes sense. What appeared to be a random set of combinations, perhaps even a different set for each respondent, all combinations, all vignettes constructed by design quickly gives way by regression analysis to a clear picture of what specific messages ‘drive’ the response. The data are often clear even with the whole panel whose individuals may possess different points of view. The pattern emerges. When the data from the panelists is analyzed by clustering, the different points of view end up being separated into mind-sets, clusters of people showing different patterns of elements which drive the responses. The foregoing has presented a short history of a growing science, beginning in the world of food [7], beginning with commercial issues, soon morphing inexorably to issues of social importance, whether that be society, ethics and morals, just to name a few [8]. Over time Mind Genomics thinking further expanded its reach into the law [9], education [10] and medicine [11] continuing to do so today, as will be shown by the study discussed here.

The Introduction of AI to the World of Systematic Experimental Design of Ideas

The migration of research away from simple surveys to the evaluation of mixtures of messages was introduced to th research world in 1993, at the annual ESOMAR research congress, held in Copenhagen, Denmark [12]. The interest by market research companies was awakened, and a number of companies licensed the technology, which was run on a PC, and required local set up, but had automatic analysis. The one continuing issue in the licensing business was the need develop the messages. Although the researchers were familiar with qualitative research (REF), the task of developing meaningful messages ‘on demand’ and for different topics began to stress the researchers, and in turn their clients. The requirement to develop a specific set of messages for a topic, messages with substance, proved harder to fulfill than simply creating a questionnaire with general ideas (e.g. how important is XYZ?) The problem of developing elements to put into the system continued to plague the growth of Mind Genomics, even though with experience most practitioners felt comfortable developing the requisite set of messaging. Over time the Mind Genomics system developed into a templated process. The user was instructed to assign the study a name, to think of four questions which tell a story, and then for each question to think of four answers. Figure 1, Panel A shows the request to provide four questions which tell a story. Creating a template helped the process of coming up with messages, although users of the program, www.BimiLeap.com, often reported that being confronted with the empty template and a request to provide questions was unnerving. Once the user developed the questions, however, providing answers became far less stressful. It became increasingly clear that the ability to think in a creative fashion seemed to have diminished, at least for the development of questions which ‘told a story.’ At the same time, it became obvious that people with little experience rarely complained about providing answers to questions that they had developed or that were posed to them by someone else. The introduction of artificial intelligence in 2022 through Chat GPT provided a breakthrough. The use of AI had been already demonstrated several years before [13]. The task was detailed, and not sufficiently quick, certainly not the ‘turnkey’ process, although for its time the process proved excellent. What was need was a turnkey system. Chat GPT by Open AI [14] provided that system, one that was incorporated into the BimiLeap platform, which supported the now more mature science of Mind Genomics. AI was introduced in the form of a tool called Idea Coach, which required the user to write out the problem (Figure 1, Panel B), which would then result in the return of 15 questions. The process could be repeated, the expression of the problem could be edited, and the process could be repeated ad infinitum. Idea Coach allowed the user to select questions, populate the template, and then do the same thing to create each of the four answers, to each o the four questions. In other words, the Idea Coach embedding AI made the process faster.

FIG 1

Figure 1: The templated request for four questions (Panel A), and the box where the user can communicate the request to AI embedded in the Idea Coach (Panel B).

Moving Beyond Questions and Answers to AI as a Deeper Guide to Critical Thinking

The initial use of Idea Coach focused on creating questions, then answers to the questions, and finally doing the ‘experiment’ with real people after having selected the four questions, and the 16 answers (elements). Over time, however, the appeal of AI grew, along with the unexpected discovery that AI would generate more when the squib (Figure 1, Panel B) was given more detail instructions, and asked to play a greater role. Table 1 shows what happens when the request for questions is replaced by a composition about a topic, specifically the issue of the Texas border. The input to the AI is far deeper, far more extensive. Table 1 shows a virtual briefing, requiring AI to carry out specific actions and exhibit critical thinking. The initial experience with this approach simply requested AI to identify mind-sets, rather than simply to prepare a set of questions. Some months later the request for AI was far deeper, with many more levels. The ability to request answers with ‘critical thinking’ became a newly discovered benefit in the Mind Genomics platform. It was easy to put together these ‘squibs’ to put into the Idea Coach. Once the squib was created the user needed only to request one iteration after another, with the AI in Idea Coach returning with either new results, or occasionally excuses why the request could not be fulfilled anew, even though the request had just been fulfilled three times, each time with different types of answers .As a second benefit, it was quite easy to switch to an edit mode, revise the squib and try again, a capability which led to a half a dozen to two dozen iterations and edits in a matter of 30 minutes.

Table 1: The request to AI to provide ‘critical thinking’ about the border issue between Texas the US Federal Government.

TAB 1

Table 2 shows the unexpected outcome. The request to the AI was to identify 18 different mind-sets, and to answer four questions (name of mind-set; believed ‘right action’, support for belief; numerical estimate of the strength of case). No other information was provided to the AI in Idea Coach. Yet, with 20 seconds of the request, the following set of 16 mind-sets (not the requested 18) was provided. The mind-sets may or may not be accurate. What is impressive, however, is both the speed at which this request was fulfilled, and the quite deep nature of the request, including the totally unexpected assignment of scores for the strength of the, or here really the ‘strength of the opinion in a court of law.’ Whether or not these conjectured mind-sets are correct is not relevant for right now. The important thing to keep in mind is that the AI allows the user to begin to think in a new manner. Rather than stranding the user with questions which require the user to enter into a potentially stressful situations, viz., required to produce ideas, the Idea Coach produces material which engages the user in a dialog. The user can discuss the conjectures produced by the AI. The AI is becoming a tutor. What is even more interesting is the ‘seeming’ ability of AI to produce estimates of the strength of the different arguments held by the mind-sets, as shown by the rating of the ‘Strength of Texas’ case’, ranging from a high of 90% (Mind-Set #6; The State’s Rights Advocate) to a low of 15% (Mind-Set #16; The Free Movement Advocate). The reasons for the ratings of strength of argument are not know, but the numbers themselves can serve as springboards for discussion.

Table 2: The 16 returned mid-sets, showing the name of the mind-set, information about the attitudes of the mind-set, and the hypothesized strength of the Texas argument, as estimated by the AI tool.

TAB 2(1)

TAB 2(2)

The Idea Coach has been programmed to provide a detailed summarization of the information at the end of the iteration. Once the steps are completed which create the ‘questions’ and ‘answer’ (viz., silos or categories and their elements) the Mind Genomics platform assembles all the iterations, and for each iteration summarizes that iteration using a series of questions. The summarized material, including that shown in Tables 1 and 2, is returned by email in an Excel file. Each tab of the Excel file corresponds to a different iteration.

First Set of Summarizations – Viewpoints and Audiences

Summarization 1 is alterative viewpoints. For this effort, the Idea Coach returned with five general viewpoints, looking at the problem of migration from a variety of points of view. The Alternative viewpoints are general in nature. Once again the alternative viewpoints are presented in descending order of strength for Texas’ case. Summarization 2 describes ten different groups who would be interested in the topic, and why specifically they would be interested in the topic. Summarization 3 describes six different groups who would be opposed to dealing with the topic for their own specific reasons. Once again it is obvious that the AI produces seemingly ‘reasonable’, ‘thought-out’ analyses. Even if the analyses are incorrect, they serve as a springboard for discussion. Once again it is important to keep in min that these summarizations are done quickly, inexpensively, and can be done at an ‘industrial-scale’ to produces specific ‘knowledge’ on virtually any topic (Table 3).

Table 3: Three different ways of approaching the issues, by considering the topic from the mind of the person. These three ways are alternative viewpoints (general viewpoints), interested audiences (the topic is interesting to discuss), and opposing audiences (the topic is not of interest to them, and they would rather see the topic ‘shelved’ or otherwise eliminated.)

TAB 3

Second Set of Summarizations – Ideas, Themes, Perspectives

This second set of summarizations focuses on the different ideas which emerge from the AI deep thinking, or deep simulation. The first summarization deals with the key ideas emerging from the exercise. The AI embedded in the Idea Coach takes the set of suggestions (questions, proposal of mind-sets), analyzes the text, and returns with the different basic ideas. In this way the summarization pulls together the different points into what could be a smaller set of more general ideas. The second summarization, themes, further reduces the information almost to a summary of the summary. The third summarization, perspectives, moves from reducing the ideas to expanding the ideas into points that can be discussed. At this point the AI has taken the user through ideas, through reducing the ideas to a limited set, and then having digested the information returns with points for further discussion (Table 4).

Table 4: Three different ways of sharpening the ideas. Key Ideas show the topics extracted from the different mindsets. Themes further reduce the information to a small set of general ideas. Perspectives challenge the reader to rethink the topic with greater insight, presenting arguments, both the pros and the cons.

TAB 4

Third Set of Summarizations – What is Missing, and Suggestion for Innovation

The objective of this third set of summarizations revolves around using Mind Genomics and Idea Coach to drive innovation. The reality of having AI suggest both what is missing, and design innovations becomes the capstone of the effort. The sheer richness of information in Table 5 is remarkable, especially recognizing that the AI embedded in Idea Coach was given no starting information. Rather, it is simple the repeated summarizations emerging from one ‘briefing’ to the AI (Table 1) which generated this set of suggestions for innovation. Of course, the suggestions must be vetted for feasibility, and tested for acceptance by people.

Table 5: The search for ‘what is next’. The top of Table 5 shows what’s missing. The bottom shows the AI suggestion for innovative ideas.

TAB 5

Discussion and Conclusions

The project reported here represents what one can accomplish with today’s AI, specifically the AI embedded within the Mind Genomics platform, www.bimileap.com. The simple approach of presenting a current issue and the plethora of information, suggestions, directions, and so forth becomes ever more remarkable when one realizes the simplicity of execution. The issue no longer becomes the ability to sit in a room of experts who have developed an understanding of the situation, that understanding coming after years of involvement, that understanding being the possession of the few, and of course in the world of homo economicus, that understanding coming at a high price tag. Rather, knowledge, direction, opportunity now lie at the hands of amateurs, those who are to inherit the future, rather than to profit from their experience. To end this paper, it is important to recognize that one can do the type of work here for virtually any topic where the mind of people is the driving force. The opportunity is to create a massive database of the mind, dealing with hundreds or even thousands of topics, these topics being general issues facing people, or being specific issues facing just a few people. One can only imagine what will happen to a world where anyone, anywhere can avail themselves of this technology as part of their education, even at the level of an elementary school student (see Mendoza et. al., 2023). The future in that world remains to unfold, but the tools are already here.

References

  1. Correa JG, Thomas JM (2024) “It’s my home, not a war zone”: It’s my home, not a war zone”: Mobilizing a multitude to demilitarize the Texas Rio Grande Valley. Sociology Compass 18(1): e13093.
  2. Glover A (2024) It’s my home, not a war zone”: Mobilizing a multitude to demilitarize the Texas Rio Grande Valley: Confronting Trump’s Reign of Terror, p. 379.
  3. Gonzalez V (2024) From Texas v. United States to United States v. Texas: The Increasing Influence of State Attorneys General on Federal Immigration Policy Through the Strategically Offensive Use of State Standing Doctrine. UCLA: New Perspectives.
  4. Hallett N (2024) The border’s migration. In: University of Chicago Legal Forum, Vol. 2023, No. 1, p.6.
  5. Reid D (2024) The Rise of Legislat,ive Intervention, the Fall of the Duty to Defend, and the Problems Thereinfor the State Attorneys General. UCLA: New Perspectives. Retrieved from
  6. Simmons B (2024) International Borders: Yours, Mine, and Ours. In University of Chicago Legal Forum 2023, No. 1, p. 10.
  7. Moskowitz HR (2012) ‘Mind genomics’: The experimental, inductive science of the ordinary, and its application to aspects of food and feeding. Physiology & Behavior 107: 606-613. [crossref]
  8. Moskowitz HR, Gofman A (2007) Selling Blue Elephants: How to Make Great Products That People Want Before They Even Know They Want Them. Pearson Education.
  9. Moskowitz HR, Wren J, Papajorgji P (2020) Mind Genomics and the Law. LAP LAMBERT Academic Publishing.
  10. Mendoza CL, Mendoza CI, Rappaport S, Deitel J, Moskowitz H, et al. (2023) Empowering young people to become researchers: What do people think about the different factors involved when shopping for food? Nutrition Research and Food Science Journal 1: 1-9.
  11. Gabay G, Gere A, Zemel G, Moskowitz H (2022) Personalized communication with patients at the emergency department—An experimental design study. Journal of Personalized Medicine 12: 1542. [crossref]
  12. Moskowitz HR, Martin D (1993) How computer aided design and presentation of concepts speeds up the product development process. In ESOMAR Marketing Research Congress (pp. 405-405). ESOMAR.
  13. Zemel R, Choudhuri SG, Gere A, Upreti H, Deitel Y, Papajorgji P, Moskowitz H, et al. (2020) Mind, consumers, and dairy: Applying artificial intelligence, Mind Genomics, and predictive viewpoint typing. In: Current Issues and Challenges in the Dairy Industry 2019 May 24 (ed: S.A. Ibrahim, T. Zimmerman, R. Gyawali). Intech Open IBSN: 9781789843569,
  14. Open AI (2023) https://beta.openai.com/docs/models/gpt-3, accessed January 18, 2023.
FIG 1

Mind Genomics, Artificial Intelligence and Today’s Realpolitik: The Disparity between the Rich and the Poor as Manifested by Mind-sets and by Slogans for the Upcoming 2024 Presidential Election

Leave a reply

DOI: 10.31038/JIAD.2024111

Abstract

A description of today’s financial disparities between the rich and the poor was fed to Idea Coach (AI) in the Mind Genomics platform (www.bimileap). Idea Coach was instructed to generate eight different ‘mind-sets’ of people based upon the description of the disparities, and to answer a variety of questions about these mind-sets. The eight mind-sets and their AI-generated features were fed to a second run of Idea Coach to generate slogans for a presidential campaign, assuming different factors, such as the year of the campaign, the nature of the person who would be the recipient of the slogan, etc. The exercise revealed the potential for deeper understanding of social issues, as well as the prospect of asking what-if questions in real time, low cost, for any topic that can be explicated simply and directly. The exercise further revealed the ability of AI to synthesize additional insights, returning all of the information in an Excel-formatted, easy-to-read ‘Idea Book.’ The process takes minutes for each iteration with the Idea Book combining all of the iterations from one run and returned to the user within a half hour.

Introduction

Economic disparities among people is a fact of life, and indeed often lies at the heart of major changes in society. The origin of this paper came from a life-long interest by the senior author, AK, in the seismic changes of society due to these economic inequalities, especially the inequality emerging today in the United States [1]. The popular term for the ultra-rich is the ‘1%’ [2]. Again and again, as the concentration of wealth increases, we end up seeing the somewhat discouraging charts showing how truly wealthy the 1% or more realistically the simply rich and ultra rich own, compared to everyone else. And, to make things worse, the stories become ever more heartbreaking as the conspicuous consumption of the ultra-rich is flaunted in the face of other people in lower socio-economic strata who have seen their lives damaged by inflation, their hopes for better economic lives buried under the insistent reduction of what they can afford and even what they can realistically attain. With this mournful issue stated, the notion was to understand the potential of artificial intelligence to help understand the mind-sets of people. Previous work by Moskowitz et. al., on the ‘fraying of America,’ looking at the responses of American a decade ago to different aspects of society suggested that people were not of one mind-set, but rather people differed among themselves in how they felt about the economic disparities. A decade later, with the foundation laid by the earlier work, two of the authors (AK and HM) decided to apply new tools afforded by AI to the mind of people presented with the story of the United States and its discouraging economic disparities.

The paper presented here is based upon the use of AI in an emerging science known as Mind Genomics [3,4]. Mind Genomics emerged as a way to understand the way people make decisions about the everyday. Rather than looking for general principles of behavior, principles which required artificial situations manipulated by an experimenter, Mind Genomics attempted to understand decision making by presenting people with vignettes, combinations of messages presenting ordinary acts of behavior in a topic area, such as voting (REF), or banking (REF), or even eating meals in a restaurant [5]. The output of these simple experiments was the ability to put numbers of different aspects of behavior to show how these aspects drove decisions, and then divide actual people into new to the world groups, mind-sets, based upon the similarities of the patterns of behavior that they evidenced for this simple, everyday task. These mind-sets often transcended countries, although the distribution of the same mind-set differed from country to country [6]. The discovery of mind-sets became a focus of many Mind Genomics studies. The approach was easy, the respondents had no problem evaluating different combinations of messages, and from the pattern of their ratings statistics had no problems first identifying the driving powers of the different messages, and then discovering the existence of meaningful mind-sets, groups of people with different ways of responding to the same messages.

The Introduction of AI to Mind Genomic Using Idea Coach, and Its Expansion to Scenarios

The underlying objective of the Mind Genomics process was that the user would achieve more knowledge by working at the level of the granular, rather than asking high level questions. The world of consumer research was overrun with surveys, most of which wanted top-line opinions of general topics, such as one’s opinion of the wait time in a doctor’s office, or the ease of purchasing a product, and the demeanor and sales behavior of the staff in a store. These general surveys ended up producing a great deal of score-card information about products and services, emerging indices like the Press Ganey score for medical services, and similar types of numbers [7]. What was not being provided from these scores was a deeper understanding of the experienced reality of the topic. The Mind Genomics effort emerged as a tool to understand the granular experience. All the users had to do was create combinations of messages, so called vignettes which were combinations of elements. The respondent would be exposed the vignettes, evaluate each vignette, and from the evaluation statistics such as regression and clustering would end up showing the strength of the elements (driving force) and mind-sets, as was noted above. There was only one issue which stopped Mind Genomics in its tracks in many cases. That was the fact that many prospective users ‘froze’ at the requirement to create questions which tell a story, and from those questions create four answers to each question. Answers were fairly easy to create, but questions were another thing entirely. Many users froze at the prospect of coming up with these questions, not realizing that the entire purpose of the questions was to create a structure bye which anyone could develop the necessary messages or elements to be used. By forcing the user to create the questions first, it was assumed that the user would be able to create the story through the questions, and then have little trouble providing the answers. The reality was that the prospect of having to ask questions and then to provide answers to those questions turned out to be more intimidating than one might have thought. The solution was not the extensive ‘training’ to teach people how to think, an effort which occasionally worked but more often than not ended in dismal failure. Rather, the solution emerged with the popular AI program, ChatGPT [8]. The Mind Genomics program, www.bimileap.com, was outfitted with the Idea Coach. Users could write in the topic, and in 15 seconds or so receive suggestions in the form of 15 questions. The process could be iterated ad infinitum, each time with AI returning its set of 15 questions, many new, some repeats. A further benefit was that the user could simply press a button to iterate to the next effort to create 15 questions, or if desired edit the prompt, called a ‘squib’ by Mind Genomics, the request, and see what emerged with the modified input [9]. Figure 1 shows the process. Panel A shows the request for four questions, the step which intimidated. Panel B shows the screen shot with Idea Coach read to receive input from the user. Panel C shows the input provided to Idea Coach by the user. Finally, Panel D shows the first part of the output provided by Idea Coach. Scrolling down revealed the full set of 15 suggested questions. The user simply needed to select a question and the question was automatically added to Panel A. the user could edit the question as well The process was rapid, simple, low cost, creating in its wake a simple system to run these Mind Genomics studies. As a benefit, the program kept a record of these iterations, and when the study set up was complete, the program sent the user the results of iterations, along with AI summarization of the results of the iterations. The process thus became a combination of problem solving and education, one available to every researcher.

Figure 1 shows the set-up process powered by AI embedded in Idea Coach. Panel A shows the request for four questions. Panel B shows the Idea Coach screen, with the box ready for input. The input could be far bigger than shown by the box. Panel C shows the same Idea Coach screen, this time with the box filled with input. Panel D shows the part of the output. The bottom of Panel D shows the steps that the suer can take. The top bar shows the, the opportunity to select a question ad drop it into the set of questions (as well as edit in on the fly). The middle bar show the opportunity to repeat the action without making any changes, which returned the new suggestions in 15-30 seconds. The bottom bar brings the user back to the Idea Coach input, allowing the user to edit the idea Coach input ‘on the fly,’ and then instantly rerun the request.

FIG 1

Figure 1: The Mind Genomics set-up screens under control of the user, with the help of Idea Coach, powered by AI.

Putting AI in Idea Coach to the Test: Providing Situations and Requesting Considered ‘Thinking’

The remainder of this paper focuses on the expansion of AI in the Mind Genomics process, not so much to provide questions and answers for study with people as with the desire to see whether Idea Coach could act more completely, almost as a person. The notion of working with AI to study scenarios is one of the attractions of AI. The objective of Mind Genomics was and remains to weave a coherent story about how people think by presenting them with combination of ideas, and measure how strong each of the ideas was to the person participating in the Mind Genomics study. The fortuitous event leading to this paper and to its companion effort was the it was possible to move Idea Coach into a higher level of functioning, simply by presenting the Idea Coach with a structured request. For this study, the nature of the information leading to request is shown in the top half of Table 1, the section titled: What is posited to be the case today (either by specific statements, or by general attitudes). The set of statements were developed by Arthur Kover, the senior author, as part of his analysis of the current situation in the United States. The actual request itself made to Idea Coach appears in the lower half of Table 2, the section entitled What AI is instructed to do based upon assumption of eight mind-sets. The important thing to keep in mind is the extensive set-up information at the start, and the far more structured request to AI as the follow on.

Table 1: Input to Idea Coach. Top panel = statement of current conditions, presented by user to AI as ‘facts. Bottom panel = request to AI from user.

TAB 1

Table 2: The eight mind-sets generated by AI in Idea Coach, based upon the information fed to it by the user (see Table 1, Top panel).

TAB 2(1)

TAB 2(2)

Positing the Mind-sets and Discussing the Way They Think

The first outcome from the AI embedded in the Idea Coach is the list of the mind-sets. Table 2 shows the list, along with the answers to the eight questions. AI provided full answers for seven of the eight ‘suggested mind-sets.’ AI did not provide anything about the eighth mind-set, other than the name ‘alternative community.’ The important things to understand from Table 2 is the depth of information and presumed insight proffered by the AI, when put on a specific question. To be sure, the orientation in Table 1 suggested the number eight for these mind-sets in the instructions to AI. Yet, one cannot fail to be impressed that the AI ‘fleshed out’ these simple statements, providing a context that would be accepted by a researcher, a news reporter, and even a novelist. The information provided by AI, or perhaps ‘spun up’ by AI seems realistic, internally consistent, and quite similar to what a person would say. The implications of this first part of the foray into the topic of today’s issue, the disparity between rich and poor, is the potential of creating a similar set of mind-sets for the dozens or even hundreds of topics facing society, first doing the work as a general topic, and then perhaps adding additional specifications, such as region of the world, period of history, and so forth. Those advances have to wait a little longer, although the speed of AI to answer these questions make it possible to create this large database of expected mind in a reasonably short period of time. Beyond the eight names of mind-sets is the ability of AI to provide answers which seem reasonable at first glance. Sometimes the AI failed to do exactly what the request to Idea Coach specified, such as giving the precise number of sentences or ideas requested, but for the most part the AI delivered what the user requested. Once again, the language and meaning of what was delivered makes intuitive sense, although successive iterations produced different words for the same questions. The variation from iteration to iteration suggests that the AI was using different sets of materials each time, although most likely materials from the same general set. Were there to be enough time and interest, one could run the same request 10 or more time, to see just how much variation in language and tonality would emerge across the 10 iterations. That is a potential topic for future work, to get a sense of the range of material and meanings delivered by AI across different efforts albeit for identical question.

After the user has completed the study set-up the Mind Genomics platform, www.bimileap, returns with summarizations of the idea which had emerged from the initial efforts, viz., questions and answers appearing in Table 1 (background materials), and Table 2 (questions and answers emerging from AI). In a sense Table 3 and successive ‘summarization’ tables provide new knowledge developed by AI.

Table 3 shows three different summaries:

  1. Key ideas from the material provided
  2. Themes emerging from the materials generated by AI
  3. Perspectives on the themes (viz., commentary or new knowledge emerging from further analysis of the themes)

It becomes increasingly clear from this table that the incorporation of AI analyses into the project moves the statement of current conditions (Top panel of Table 1) into a far more profound ‘exegesis’ of the topic. Once again, Table 3 provides the input to help critical thinking about the topic.

Table 3: Summarization of key ideas, themes, and perspectives regarding the themes

TAB 3

 

The next step is technically not a summarization of the material, but rather the question of who would be positive about this material (Interested Audiences) versus who would be negative about this material (Opposing Audiences). Table 4 shows these two audiences.

Table 4: Relevant aspects of the issue for Interested Audiences versus for Opposing Audiences

TAB 4

 

A key benefit of the AI embedded in Idea Coach comes from the ability to identify new aspects, hitherto either unknown or perhaps not particularly well identified. Table 5 shows the final set of summaries. The first summary shows Alternative Viewpoints, which comprise a set of 15 questions presenting new ways of thinking about the topic. The second summary shows What’s Missing, comprising 12 direct questions. The third summary shows suggestions for innovations. This third summary on innovation looks at the topics, recasts the issue in terms new mind-sets for each new idea, and then presents the innovation, and how it will affect each of the newly minted mind-sets. Once again the summarization is new knowledge or at least conjecture, created by AI.

Table 5: New perspectives and suggestions emerging from AI

TAB 5

Creating Relevant Slogans for the Presidential Election Using the Mind-sets Generated by AI in Idea Coach

The final product of this experiment appears in Table 6. The product, in the second panel, comprises one slogan for each of the eight mind-sets. Each column corresponds to one mind-set. Each row corresponds to a hypothetical period of time (e.g., 2024, 1892, 1860) to a hypothesized person (e.g., Total US citizen, registered Democrat, registered Republican, registered Independent), and to a statement of economic condition.

Table 6: Slogans developed by AI, based upon instructions to Idea Coach

TAB 6

Discussion and Conclusions

The original thrust of this paper was to explore whether or not artificial intelligence could be used in the world of social policy [10,11], especially to enhance creativity in an area where opinions and power dominate [12]. The topic of AI and social policy has become exceptionally complex, often with the explorations shown in this paper overshadowed by the ethical issues [13,14]. The sheer fact of incorporating machines into society in a way which has the flavor of ‘control’ brings up the shades of George Orwell’s 1984 [15], and predecessor writings of this ilk [16]. The discussions of AI and ‘social research’ focuses primarily on ethics rather than on advances [17,18]. The range and depth of the discussions on AI do not so much focus on specifics but rather on discomfort, fears, which end up showing themselves as seemingly never-end discussions of ethical implications at the level of generality, rather than of specific issues. It may well be that the real contribution of AI to society and to governance have yet to be made, in contrast to the contribution of AI to applications, such as medicine [19], library work in the law [20], etc. It also may be the case that there is justification for the application of AI in situations of power of people over people, with AI giving an advantage in the ‘battle of all against all,’ the perfect description of the Hobbesian nightmare. Such nightmares would not be the case for the benign and often useful application of AI to solve humanity’s problems, rather than the fear of AI which, instead, could create these problems [21,22]. The paper just presented is one in a series of papers exploring the potential of artificial intelligence to promote critical thinking. The issue of critical thinking in the academic world is well covered in numerous publications [23]. There are methods for teaching critical thinking at a young age. The emphasis on critical thinking is to get the student to delve more deeply into a topic, develop hypotheses, test these hypotheses by one or another fashion [24,25]. One of the continuing issues in today’s world is how to make students focus on topics, especially in a world where their attention wanders from the topic or teacher to the little screen, their phone or tablet, where they can be entertained. Is it possible to teach critical thinking within that environment? The study or simulation study reported here may be one way to teach critical thinking about a topic. The issue here is positing and understanding mind-sets, and then creating slogans. The topic is serious, but at another level it is fun, not particularly challenging, returns with interesting results, and most important, returns with results which are real, and which themselves make a major contribution. It would be premature as well as difficult to enumerate the applications of the approach presented here. Within just a few minutes one could take a topic, provide one’s information of any type, instruct the AI to assumed mind-sets without saying what these mind-sets are, and then instruct the mind-sets to answer specific questions. At the surface level the answers make intuitive sense, perhaps because they are not presents facts but trends. It may be at this level, trends, generalities which make intuitive sense, that the approach presented here may enjoy its earliest success. Be that as it may, there may also be a way of presenting the AI with the right question so that the process takes place sometime in the well-defined past, or even sometime in the future. In those cases, one would assume that the AI uses the available information to extrapolate to the new condition. That extension remains for future work, but the effort, cost, speed, and simplicity all argue for an interest ‘next step’ along those lines. The ingoing vision of this study was to understand AI and politics, especially elections. The exiting vision is of a tool in the hands of billions of students, all of whom have fun playing with a system which educates them, makes them prospective experts in a topic with very little difficulty, almost overnight, and for very low cost. To close this paper, it is use to think of the thrust of which this paper is simply the latest. The paper is one of a series of papers exploring the potential of artificial intelligence to promote critical thinking. The issue of critical thinking in the academic world is well covered in numerous publications. The emphasis on critical thinking is to get the student to delve more deeply into a topic, develop hypotheses, test these hypotheses by one or another fashion [26,27]. In the end, it will be critical thinking, hand in hand with the increasing power of AI, which will see the proper use of technology to make a better world, not a world of machines and power-hungry individuals in control.

References

  1. Moskowitz H, Kover A, Papajorgji P (eds) (2022) Applying Mind Genomics to Social Sciences. IGI Global.
  2. Harrington B (2016) Capital Without Borders: Wealth Managers and the One Percent. Harvard University Press.
  3. Moskowitz HR (2012) ‘Mind Genomics’: the experimental, inductive science of the ordinary, and its application to aspects of food and feeding. Physiology & Behavior 107: 606-613.
  4. Moskowitz HR, Gofman A, Beckley J, Ashman H (2006) Founding a new science: Mind genomics. Journal of Sensory Studies 21: 266-307.
  5. Mazzio J, Davidov S, Moskowitz H (2020) Understanding the algebra of the restaurant patron: A cartography using cognitive economics and Mind Genomics. Nutrition Research and Food Science Journal 3: 1-11.
  6. Rabino S, Moskowitz H, Katz R, et al. (2007) Creating databases from cross‐national comparisons of food mind‐sets. Journal of Sensory Studies 22: 550-586.
  7. Siegrist Jr RB (2013) Patient satisfaction: history, myths, and misperceptions. AMA Journal of Ethics 15: 982-987.
  8. Abdul ah M, Madain A, Jararweh Y (2022) ChatGPT: Fundamentals, applications and social impacts. In 2022 Ninth International Conference on Social Networks Analysis, Management and Security (SNAMS) IEEE, pp 1-8.
  9. Moskowitz H, Todri A, Papajorgji P, et al. (2023) Sourcing and vetting ideas for sustainability in the retail supply chain: The contribution of artificial Intelligence coupled with Mind Genomics. International Journal of Food System Dynamics 14: 367-380.
  10. Cheng L, Varshney KR, Liu H (2021) Socially responsible ai algorithms: Issues, purposes, and challenges. Journal of Artificial Intelligence Research 71: 1137-1181.
  11. Schiff D, Biddle J, Borenstein J, Laas K (2020) What’s next for ai ethics, policy, and governance? a global overview. In Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society, pp. 153-158.
  12. Gobet F, Sala G (2019) How artificial intelligence can help us understand human creativity. Frontiers in Psychology, 19 June 2019 Sec. Cognition Volume 10 – 2019. [crossref]
  13. Oravec JA (2019) Artificial intelligence, automation, and social welfare: Some ethical and historical perspectives on technological overstatement and hyperbole. Ethics and Social Welfare 13: 18-32.
  14. Ouchchy L, Coin A, Dubljević V (2020) AI in the headlines: the portrayal of the ethical issues of artificial intelligence in the media. AI & SOCIETY 35: 927-936.
  15. Sudmann A, Waibel A (2019) That is a 1984 Orwellian future at our doorstep, right? Natural Language Processing, Artificial Neural Networks and the Politics of (Democratizing) AI.
  16. Roazen Paul (1978) The Virginia Quarterly Review 54: 675-695.
  17. Floridi L, Cowls J, King TC, d Taddeo M (2021) How to design AI for social good: seven essential factors. Ethics, Governance, and Policies in Artificial Intelligence, pp. 125-151.
  18. Hong JW (2022) With great power comes great responsibility: inquiry into the social roles and the power dynamics in human-AI interactions. Journal of Control and Decision 9: 347-354.
  19. Subramanian M, Wojtusciszyn A, Favre L, et al. (2020) Precision medicine in the era of artificial intelligence: implications in chronic disease management. Journal of Translational Medicine 18: 1-12.
  20. Armour J, Sako M (2020) AI-enabled business models in legal services: from traditional law firms to next-generation law companies?. Journal of Professions and Organization 7: 27-46.
  21. Assibong PA, Wogu IAP, Sholarin MA, et al. (2020) The politics of artificial intelligence behaviour and human rights violation issues in the 2016 US presidential elections: An appraisal. In Data Management, Analytics and Innovation: Proceedings of ICDMAI 2019, Volume 2 (Springer Singapor, pp. 295-309).
  22. König PD, Wenzelburger G (2020) Opportunity for renewal or disruptive force? How artificial intelligence alters democratic politics. Government Information Quarterly 37:101489.
  23. Kreps S, McCain RM, Brundage M (2022) All the news that’s fit to fabricate: AI-generated text as a tool of media misinformation. Journal of Experimental Political Science 9: 104-117.
  24. McPeck JE (2016) Critical Thinking and Education. Routledge.
  25. Jenkins EK (1998) The significant role of critical thinking in predicting auditing students’ performance. Journal of Education for Business 73: 274-279.
  26. McLaren BM, Scheuer O, Mikšátko J (2010) Supporting collaborative learning and e-discussions using artificial intelligence techniques. International Journal of Artificial Intelligence in Education 20: 1-46.
  27. Westermann C, Gupta T (2023) Turning queries into questions: For a plurality of perspectives in the age of AI and other frameworks with limited (mind) sets. Technoetic Arts: A Journal of Speculative Research 21: 3-13.
fig 1

Mind Genomics and Today’s Realpolitik: Considering the ‘Invasion’ of Single Young Men at the US Southern Border from the Point of View of What Mind-sets Might Exist and What to Consider

Leave a reply

DOI: 10.31038/MGSPE.2024414

Abstract

A combination of Mind Genomics to understand motivation coupled with Idea Coach (artificial intelligence module within Mind Genomics) was used to create synthetic mind-sets which might describe young males illegally crossing the US southern border. The paper shows how AI can provide information to spur critical thinking when provided with a description of the situation and the motivation for the illegal crossing. The authors suggest that the world of law enforcement might benefit by using these procedures to facilitate critical thinking.

Introduction

As of this writing, February 2024, the United States is experiencing a never-before-situation at its borders. The southern borders of the United States, especially those in Texas, are being inundated by migrants, many of whom are young, unaccompanied men, who slip into the United and end up disappearing inside the United States. Many of these people disappear entirely. Others go to court and are allowed to say pending their case.

Up to now, the material just presented recognizes a problem emerging, namely the escape of many unregistered aliens through what has turned out to be an exceptionally porous border, manned by seriously understaffed border patrols and immigration officers. The consequence is not unexpected. Many people believe that through such unmonitored immigration, there is a good chance at an alien army may be coming in, this army not necessary friends to the United States, the company into which they are disappearing.

The objective of this paper is to demonstrate how AI can be used to formulate hypotheses about the nature of what might be happening at the southern border, and then to demonstrate the change in AI-based ‘conclusions’ when the motive for the invasion includes intending to harm the United States.

History of the Approach

The tool used here is Idea Coach, an AI-empowered program embedded in the Mind Genomics platform [1]. In turn, Mind Genomics is a platform which specializes in the analysis of human judgments, doing so by presenting the respondent (survey taker) with a variety of vignettes, and for each vignette, obtaining a rating on defined scale. The vignettes themselves are combinations of simple, easy-to-read statements, called elements. The vignettes are constructed according to a plan known as an experimental design. Each respondent evaluates a unique set of 24 vignettes [2]. The ratings are then transformed to a simple Yes/No scale. The final steps are to use simple OLS (ordinary least-squares) regression at the level of each respondent to relate the presence/absence of the elements to the transformed ratings. The coefficients of the equations become the tool to understand the mind of the respondent. This simple analysis shows immediately what elements ‘drive’ the ratings, and by so doing reveal the underlying mind of the respondent with regard to the specific topic. The final analysis clusters the respondents into different groups based upon similar patterns of coefficients.

The foregoing approach has been embedded into the aforementioned Mind Genomics platform (www.bimileap.com). The original format of Mind Genomics required that the user provide a set of four questions ‘which tell a story’, and then ‘four answers to each question.’ It was the answers which the respondent evaluated, after these answers (now called elements) were mixed and matched to create the vignette.

AI was necessary to help the user think of questions and answers. Over the period of several years, it became increasingly obvious that users of Mind Genomics both liked the approach but were terrified of the requirement to come up with questions and answers. Figure 1 shows the request to provide four questions. This request was a wall to many prospective users because, quite simply, it was daunting. People were often good at answering questions but not at formulating questions to tell a story.

The incorporation of AI into the process of questions and answers increased the acceptance of the Mind Genomics platform, for at least four reasons:

  1. The process no longer stymied the user. A simple ‘squib’ in Figure 1, Panel B, sufficed to generate 15 questions.
  2. The same process occurred in the generation of elements. Once the user selected four questions and put those questions into the template (Figure 1, Panel A), the Idea Coach was able to return 15 answers to each question selected.
  3. The process was rapid, with the suggested sets of 15 questions or 15 answers to a question returning in about 20 seconds.
  4. The user could edit the squib to change the nature of the questions, or edit the selected questions to change the nature of the answers.

fig 1

Figure 1: The set-up screen for Mind Genomics studies. Panel A requires the user to provide four questions which tell a story. Panel B shows the Idea Coach, giving the user the opportunity to describe the topic, and turn receive 15 questions generated by AI.

The incorporation of AI as Idea Coach ended up producing Idea Books, compilations of questions produced in response to the squib (Figure 1, Panel B), as well as compilations of answers produced in response to each question. Each page in the Idea Book corresponded to one iteration, whether the 15 suggested questions resulting from a request written out in the squib, or 15 answers resulting from the selection of a question. It was not unusual to generate Idea Books of 10+ tabs.

In addition to the sets of questions or answers on each page, the AI was given the task of summarizing the material on each page, viz., the questions or answers. The result was other insights, such as key ideas, themes, perspective, what is missing, interested audiences, opposing audiences, and innovations. Each of the foregoing was given its own section on the page in bold type, and then the relevant AI summarization provided.

The foregoing process required about 30 minutes in total from beginning the set up of the Mind Genomics experiment to the creation of a book with say 20 pages. The process itself was quick, the results were easy to obtain, and the iterations themselves became a source of learning, the Idea Book turning into a resource book for further work.

Over time, and as the process became easier, the process first became rigid as practitioners using Mind Genomics followed the path laid out, with simple questions posed to the AI embedded in Idea Coach. The only modifications during the early days of Idea Coach, the year 2022-2023, was the expansions of the question, so that the questions would have a certain number of word (~ 10-15), that the questions would be interesting, and that the questions could be understood by a young person. The same ‘editing’ of requests was done for the questions themselves in order to generate answers which ‘were not lists, but rather statements which could lead to a discussion’. All of these were style questions, rather than substance questions. The happy consequence was that the Idea Book was richer in content, the questions and answers more instructive, and the process enjoyable to the user, who could practice writing different requests about the style and structure of the output to be generated by Idea Coach

Advancing Insights Through Deeper Interactions with AI Through the Idea Coach

During the course of working with Idea Coach, author Mulvey expanded the nature of the squib, and generated unexpected and deeper results. Rather than simply specifying the nature of the question or answer in terms of style (viz., number of words, age of reader, style to engage the reader rather than list options), Mulvey added a request into the squib. That request was to provide some additional answers to the question. That is, the squib or input to the AI embedded in Idea Coach contained a request for additional structure in the question, rather than just a question alone. The approach is conceptually similar to the creation of synthetic data, in this case synthetic mind-sets [3].

Idea Coach returned with unexpectedly deeper questions. The output to this ‘expanded request’ was more like a summary of a situation than simply suggested questions. Exploration of alternative ways to expanded the input to Idea Coach quickly revealed that the AI could be requested to a far deeper analysis. And so the approach was born which lies at the foundation of this paper.

The paper itself explores how the Idea Coach provides more information when given a detailed instructions. Table 1 shows the text as it appears in the Idea Book returned to the respondent. The top section in bold shows what the user types into the squib. The rest of the table shows what it returned to the user by Idea Coach when the Idea Book is completed.

Table 1: Output from Idea Coach for the first request, where the hostile goals of the border crossing men are not revealed.

tab 1(1)

tab 1(2)

tab 1(3)

tab 1(4)

The first part of the input to Idea Coach is simple and direct, describing what is happening., and a simple request for suggestions: The topic is: Invasion of the United States at the southern border cross by unmarried young men of military age. How can we prevent these people from starting massacres at unarmed gatherings throughout the United States if they are truly ‘invading us’.

The second part of the input to Idea Coach is the set-up hypothesis that there exist three mind-sets. These mind-sets are not named, and indeed no information is given about any conjectures regarding the mind-sets. It will be the job of the AI in Idea Coach to suggest the mind-sets. We will see that the suggested mind-sets returned by AI ends up concurring with additional information provided by the user. The actual text is straightforward: We believe that there are three different mind-sets of these young men. The user can change the number of mind-sets and add more information about the mind-sets. Each change generates a new set of responses, further serving as an educational and preparedness tool for the user. and can be modified by the user to see what happens when the number of mind-sets is changed.

The third part of the input to Idea Coach is the instruction to answer a set of eight specific questions for each to-be-named mind-set.:

For each mind-set in turn, answer these specific questions

  1. What is the name of the mind-set
  2. What is the goal of the mind-set specifically with regard to being in the United States
  3. What are they likely to say to an immigration official when they are caught
  4. What will convince them to go through legal channels to become regular citizens
  5. How can we recognize them… give four indications to help recognize the mind set to which they belong
  6. If unrecognized, what will they likely do in three months after they have entered illegally into the United States
  7. If unrecognized, what will they likely do in six months after they have enter illegally into the United States
  8. What will make them want to identify themselves to officials in the United States

Do the above answers for each mind-set separately, answering all questions 1-8 for each mind-set in turn

The fourth and final instruction to Idea Coach focuses on the style of the suggestion to be given by Idea Coach:

Make the answers interesting to read, and easy to talk about to other people

Make the answers as realistic as possible

Recognize that the answers will be shared with officials in the United States

Section B in Table 1 shows the immediate work-product of the AI embedded in Idea Coach. Section B returns with 30 seconds. Generally, but not always, Idea Coach provides precise sets of answers, following the instructions written in the squib. The user is free to re-run the Idea Coach many times, changing the squib desired. (That change will be shown in Table 2, where the input to Idea Coach will add information about the migrants ‘wanting to harm America’.) Each iteration of Idea Coach generates a mix of new ideas and old ideas

Section C in Table 1 shows a set of summaries created by AI, with these summaries taking into account all of the information in Sections A and B, respectively. The summarizing queries in Idea Coach are fixed, but the information in each iteration tends to be partially unique, so that the summarizations will differ from iteration to iteration.

When looking at different sections of Table 1 the reader should keep in mind that within less than a minute the user has gone from a set of questions to a set of answers, and in a few more minutes from a set of answers to a set of summarized generalities. It is also important to keep in mind that the typical Idea Book does not stop with one or two iterations, but may go on for dozens of iterations, each done without much effort, each done to satisfy one’s curiosity about a particular issue, not only about the general topic.

Iterating and Adding Information about Motives for a More Targeted Analysis

As stated above, a key benefit of the Mind Genomics approach as empowered by the Idea Coach emerges from the ability to modify the squib or request given to the underlying AI. Table 1 showed the three mind-sets without any specification of who the mind-sets are, other than the general concern about a possible massacre in a public place.

Tables 2-5 show the descriptions of the mind-sets developed by AI when the invading males are further specified as to their assumed motives. The issue now is to see how these three mind-sets are described. For the purposes of this paper, the focus is simply on the types of descriptions which emerge when additional information is provided to the AI embedded in Idea Coach. In a sense the descriptions of the mind-sets generated in Tables 1-5 can be looked at an exploration of how AI can put features onto basic descriptions.

Table 2: Three mind-sets emerging from AI when the motivation is to start a family in the United States

tab 2

Table 3: Three mind-sets emerging from AI when the motivation is to start working and then become citizens

tab 3(1)

tab 3(2)

Table 4: Two mind-sets emerging from AI when the motivation is to get jobs and bring in their family living in poverty

tab 4(1)

tab 4(2)

Table 5: Three mind-sets emerging from AI when the motivation is to harm the United States with gang warfare

tab 5

Discussion and Conclusions

The ingoing rationale for this study was to demonstrate that a new opportunity to understand behavior has evolved from the incorporation of AI through Idea Coach into the basic thrust of Mind Genomics. The original objectives of Mind Genomics focused on understanding motivations and decision ‘rules’ for different types of people, with these rules emerging from material taken from the granular, everyday world. The development of AI made it possible to accelerate the process by producing a way to help people ask questions and create answers for that question.

The introduction of AI also made it possible to go into directions not thought of before, or if thought of, then consigned to the world of the theoretical. We are talking here about ‘what if’ questions. What if we could ascribe basic motivations to people, almost making the task which involves synthetic people, rather than real people. This paper shows what can be done by creating synthetic people, simply by telling AI that there are three mind-sets, giving some background, and then systematically varying some aspect of that background. Tables 1-5 show what happens when the user moves from no motivation stated to a variety of different motivations.

Up to now a major focus of AI in law enforcement has been to detect patterns in the transfer of money and other objects, almost the combination of Big Data and the Internet of Things [4-7]. The use of AI to construct synthetic mind-sets for law may be in progress but is not yet mainstream. On the other hand, the use of AI to construct synthetic people for surveys is beginning to become mainstream, at least in the market research community. There is every reason to assume that the use of AI to construct scenarios and synthetic people, as well as synthetic mind-sets, will become mainstream, and perhaps even significant. To the degree that the approach presented here becomes a tool for critical thinking in law and public policy, we may expect to see the approach presented here proliferate and improve thinking as well as public policy.

References

  1. Moskowitz H, Todri A, Papajorgji P, et al. (2023) Sourcing and vetting ideas for sustainability in the retail supply chain: The Contribution of artificial intelligence coupled with Mind Genomics. International Journal of Food System Dynamics 14: 367-380.
  2. Gofman A, Moskowitz H (2010) Isomorphic permuted experimental designs and their application in conjoint analysis. Journal of Sensory Studies 25: 127-145.
  3. Raghunathan TE (2021) Synthetic Data, Annual Review of Statistics and Its Application 8: 129-140.
  4. Borum R (2020) Scientific and technological advances in law enforcement intelligence analysis. Science Informed Policing 99-121.
  5. Rademacher T (2020) Artificial intelligence and law enforcement. In: Regulating Artificial Intelligence, (ed, T. Wischmeyer, T, Rademacher), Springer Link, 225-254.
  6. Watney M (2019) Law Enforcement Use of Artificial Intelligence for Domestic Security: Challenges and Risks. In Proceedings of the European Conference of Artificial Intelligence and Robotics (ECIAR), Oxford, UK, November, 341-348.
  7. Yavorsky MA, Mikheeva SV (2022) The Use of Artificial Intelligence Technologies and Big Data in Law Enforcement. In Proceedings of the International Conference Engineering Innovations and Sustainable Development, 669-675, Springer International Publishing.
fig 1

How AI Helps a Nurse Learn How to Better Communicate with a Sufferer of Frontotemporal Dementia: Exploring AI Driven by the Mind Genomics World-view

Leave a reply

DOI: 10.31038/IJNM.2024521

Abstract

Mind Genomics can be used to study language-based frontotemporal dementia (FTD) by synthesizing the cognitive and behavioral profiles of individuals. This personalized approach helps identify challenges and tailor interventions to each person’s needs. By recognizing the diversity of mind-sets within the FTD population, more effective strategies for diagnosis, treatment, and care can be developed. The paper shows how to use AI (Socrates as a Service) to synthesize a user-defined number of mind-sets, here three FTD mind-sets: Apathetic Mind-Set, Disinhibited Mind-Set and Compulsive Mind-Set. The combination of Mind Genomics and AI has the potential to revolutionize research, diagnosis, and care for FTD, leading to better outcomes and improved quality of life.

Introduction

Frontotemporal dementia (FTD) is a neurodegenerative disease which primarily affects the frontal and temporal lobes of the brain. FTD leads to changes in behavior, personality, and language skills. Common symptoms include changes in behavior, personality, and language abilities. Sufferers may exhibit apathy, social disinhibition, impulsivity, and emotional bluntness. The speech of FTD sufferers may become fluent but lack meaning, or they may struggle to find the right words. As the disease progresses, patients may develop memory loss and lose the ability to perform daily activities. It is important for nurses to understand these language changes in order to communicate with, and care for FTD patients [1-4]. Talking to people who have frontotemporal dementia is an important part of giving them good care and support. When working with these people, nurses need to be patient, kind, and understanding. It is best to speak slowly, use simple words, and give the patient a lot of time to answer. Body language, such as gestures, facial expressions, and the tone of one’s voice, can also help get the point across. To start a therapeutic relationship and look out for the patient’s well-being, one needs to build trust and a relationship with them. Patients may have trouble expressing their needs, feelings, and thoughts because their language skills are affected. It is helpful to use simple words, speak slowly and clearly, and show things to help people understand. It’s important to build trust and a relationship with patients so they feel safe and understood. Getting into a routine and being consistent with how one talks to them can help FTD patients feel less anxious and confused and improve their overall quality of life [5-8]. Artificial intelligence (AI) can help us learn a lot about how to help FTD sufferers. AI can help researchers find patterns, predict how the disease will progress, and come up with personalized treatment plans by looking at data from people who have FTD. AI can also create virtual simulations which tutor healthcare professionals as they practice talking to FTD patients, learning how to do it better. Incorporating AI into nursing may be a strong positive step in the evolution of medicine in this age of intelligent computing [9-11]. AI may be able to help nurses gain insight and information about how FTD suffers think and reason, giving the nurses a way to come up with better ways to talk to the FTD sufferer. By looking at data and putting it all together. AI can help nurses guess what might happen, allowing the nurse to modify their approach to fit the needs of each FTD sufferer. Nurses can learn more about how to care for people with FTD by using AI to look at huge amounts of data and then simulate different situations [12-14].

Learning about FTD by Using Mind Genomics Discoveries Regarding Mind-Sets

Mind Genomics is an emerging science focusing on how people perceive and react to the world of the everyday. Mind Genomics has a long history of application in the social sciences, marketing, and consumer research to better understand human behavior. Researchers can better connect with various groups of people by dividing populations according to their mindsets. Research has demonstrated that this customized method outperforms generic messaging when it comes to achieving desired responses from specific audiences. Mind Genomics works by developing questions about a topic, these questions telling a story, and then creating answers to the questions. The process takes the answers (aka elements, stand-alone phrases which paint a word picture), combining them into short vignettes, presenting the vignettes to survey takers (respondents). The ratings assigned by the respondents are then deconstructed by statistics (ordinary least squares regression). The output of regression, coefficients showing the ‘driving power’ of the elements, is subject to cluster analysis [15]. The output of the foregoing statistical journey are sets of people who think differently about the same specific topic, and thus who should be treated differently through communication. The Mind Genomics studies continue to reveal different clusters of people, different mind-sets. The vignettes were set up so that each respondent evaluated a totally different set of vignettes. The most recent design calls for four questions, and four answers to each question, but earlier versions called for four questions and nine answers to each question (60 vignettes), or six questions and six answers to each question). It was impossible to game the system. The mindsets ended up being coherent, interpretable, and often meaningful for subsequent communication. The outcome was thus the development of specific, granular knowledge about aspects of a topic, as well as the precise words to which the different mind-sets would react [16-18].

The Mind Genomics Process and the Introduction of AI to Help Coach the Users

As stated above, Mind Genomics was created to help users (e.g., researchers) discover how people think about the world of the ordinary, doing so by creating questions and then answers to those questions (elements). Whereas on the surface this requirement seems fairly easy, the reality in practice was anything but that. The reality turned into the recognition that structured thinking to do the seemingly simple task was more elusive. At first in 2015 and later the answer was extensive training. The training, however, was also an inhibitor, converting the satisfaction of learning into the pain of learning a new tool. The development of AI around 2022-2023 which generated the breakthrough, shown descriptively in Panel A as Idea Coach. Idea Coach was the link to AI. A specific program was developed to create questions and answers. This program used AI, and was called, not surprisingly, Socrates as a Service.

Figure 1 shows the process which led to the AI, and in turn to this paper.

  1. Panel A shows the request to develop four questions. The topic is communicating with and helping patients with FTD. As just noted, the response to providing the questions was eventually the creation of four questions, but with a great deal of angst, insecurity, and often simple frustration. It was at this point that many prospective users simply abandoned the process.
  2. Panel B shows the input to AI (Idea Coach). The user types the request into the box. The AI in Mind Genomics is a programmed set of queries (SCAS, Socrates as a Service). SCAS is programmed using ChatGpt 3.5 [19,20] to provide the questions, and the answers, depending upon the information provided to it. The imortant thing is that SCAS is quick, returning in 15 seconds, allows for numerous iterations so the user can get an education at the time when SCAS is used, and finally returns with additional post-use analyses, also in depth, serving as a way to increase learning.
  3. Panel C shows the output from SCAS. SCAS is programmed to provide 15 questions for each iteration. Subsequent iterations generate new sets of questions. When the user runs 10 iterations, it is likely that the result will be a set of 100+ unique questions embedded in the total of 150 questins returned by SCAS. The 100 or so questins creates an extensive reference library of questions, each one of which can be addressed.
  4. Panel D shows the selection of a random set of four questions to be entered into the templated sytem for Mind Genomics. These questions can be selected ‘as is’ from the SCAS output, or edited, and sometimes the questions can be inserted manually by the user, without any help from AI.
  5. Panel E shows the request for four answers for Question 1 of 4. The question comes directly from the first question in Panel D.
  6. Panel F shows the first eight of 15 answers for Question 1. Each iteration of SCAS answers one of the four questions selected, and generates 15 answers. As before, the answers can be edited, the request for answers can be iterated, and the question itself can be edited to shape the nature of the answers.

fig 1

Figure 1: The process of Mind Genomics. Panel A shows the request for four questions. Panel B shows the squib, viz., the background to SCAS (the programmed AI). Panel C shows a subset of the 15 questins which emerge from each iteration. Panel D shows the selected questons automatially entereed into the Mind Genomics template. Panel E shows the request for 15 answers to questin 1, with the text of the quetisn at the top of the screen shot. Panel F shows some of the 15 answers emergng from SCAS.

Moving Beyond User-generated Questions to AI Generated Questions

The introduction of AI into the Mind Genomics platform was done with the idea that the SCAS approach would be a source of additional learning. To that end, the creation of 15 questions was made a standard feature of the in Figure 1, Panel B. Any time that the user would engage SCAS through the user-provided squib, there outcome would be a set of 15 questions. As part of using the system, several times the 15 questions emerging from SCAS were ‘accidentally’ copied and then used for the next iteration of SCAS. The SCAS dutifully returned with an answer to each question. When this happy state of affairs was recognized, it was not long before the questions were ‘imported’ into SCAS, with requests to provide more than just a single answer. Table 1 shows the results of tentative steps to push SCAS to provide two answers and a ‘slogan’ for each of 12 questions which had had emerged from SCAS in the previous iteration. It is virtually impossible to detect the fact that these questions and answers all come from AI.

Table 1: Questions and answers about FTD, all generated by AI, with the requested information alone coming from the human user.

tab 1

Introducing Mind Genomics Thinking into AI by Hypothesizing the Nature of Three FTD Mind-sets

It was the discovery that the AI embedded in SCAS could do more than simply respond to questions which generated the next step. The question was what AI would do when given specific background information assumed to be ‘true,’ and then instructed to provide information to ‘flesh out’ the background information. The AI would be given specific information and specific requests. The ‘test’ began by telling AI that for FTD (language loss variety) there are three major mind-sets. We do not specify what they are, nor anything else. The only information that the AI receives is the specification of three mind-sets, followed by the instruction to answer specific questions. The actual information provided to the AI was thus minimal. The statement that there were three mind-sets for SCAS to ‘create’ the mind-sets in detail: Apathetic, Disinhibited, and Compulsive, respectively. Table 2 describes the Apathetic Mind-Set, Table 3 the Disinhibited Mind-Set, and Table 4 the Compulsive Mind-Set.

The final simulation effort appears in Table 5. Table continues the effort of answering a series of questions, moving beyond simple answers by directing SCAS to provide two answers and a memorable slogan.

Table 2: AI-synthesized characteristics of the Apathetic Mind-Set

tab 2

Table 3: AI-synthesized characteristics of the Disinhibited Apathetic Mind-Set

tab 3

Table 4: AI-synthesized characteristics of the Compulsive Mind-Set

tab 4

Table 5: Answers to direct questions posed by the user (Part 1) and then additional information ‘volunteered’ by SCAS afterwards. The table is constructed from several iterations.

tab 5(1)

tab 5(2)

Discussion and Conclusions

By employing Mind Genomics in the study of FTD, researchers can uncover valuable insights into the cognitive and behavioral profiles of individuals with the disease. This personalized approach can help identify specific challenges and tailor interventions to address the individual needs of each person living with FTD. By recognizing the diversity of mind-sets within the FTD population, we can better understand the complexities of the disease and develop more effective strategies for diagnosis, treatment, and care. Through the application of Mind Genomics in the field of FTD research, we can gain a deeper understanding of the cognitive and behavioral changes associated with the disease. By identifying distinct mind-sets within the FTD population, we can tailor interventions to address the specific challenges faced by individuals with different profiles. This personalized approach can lead to more targeted and effective care strategies to improve the quality of life for individuals living with FTD and their caregivers. In summary, Mind Genomics offers a powerful tool for understanding the diverse ways in which individuals with language-based FTD experience and navigate the world. By recognizing and synthesizing the unique mind-sets present within the FTD population, we can develop more personalized and effective interventions that address the diverse needs of individuals affected by the disease. This approach has the potential to revolutionize the way we approach research, diagnosis, and care for individuals with FTD, ultimately leading to better outcomes and improved quality of life.

Acknowledgement

Howard Moskowitz gratefully acknowledges the ongoing of help of Ms Hilda Varnum in this effort, and the contribution of Ms Arlene Gandler who provided the original stimulation to write this paper.

References

  1. Bang J, Spina S, Miller BL (2015) Frontotemporal dementia. The Lancet 386: 1672-1682.
  2. Onyike CU, Diehl-Schmid J (2013) The epidemiology of frontotemporal dementia. International Review of Psychiatry 25: 130-137. [crossref]
  3. Weder ND, Aziz R, Wilkins K, Tampi RR (2007) Frontotemporal dementias: a review. Annals of General Psychiatry 6: 1-10. [crossref]
  4. Young JJ, Lavakumar M, Tampi D, Balachandran S, Tampi RR (2018) Frontotemporal dementia: latest evidence and clinical implications. Therapeutic Advances in Psychopharmacology 8: 33-48. [crossref]
  5. Caceres BA, Frank MO, Jun J, Martelly MT, Sadarangani T, et al. (2016) Family caregivers of patients with frontotemporal dementia: An integrative review. International Journal of Nursing Studies 55: 71-84. [crossref]
  6. Schmid J, Schmidt EM, Nunnemann S, Riedl L, Kurz A, et al. (2013) Caregiver burden and needs in frontotemporal dementia. Journal of Geriatric Psychiatry and Neurology 26: 221-229. [crossref]
  7. Edberg AK, Edfors E (2008) Nursing care for people with frontal-lobe dementia–difficulties and possibilities. International Ppsychogeriatrics 20: s 361-374. [crossref]
  8. Rasmussen H, Hellzen O, Stordal E, Enmarker I (2019) Family caregivers experiences of the pre-diagnostic stage in frontotemporal dementia. Geriatric Nursing 40: 246-251. [crossref]
  9. Li R, Wang X, Lawler K, Garg S, Bai Q, Alty J (2022) Applications of artificial intelligence to aid early detection of dementia: A scoping review on current capabilities and future directions. Journal of Biomedical Informatics 127. [crossref]
  10. Merkin A, Krishnamurthi R, Medvedev ON (2022) Machine learning, artificial intelligence and the prediction of dementia. Current Opinion in Psychiatry 35: 123-129. [crossref]
  11. Ranson JM, Bucholc M, Lyall D, Newby D, Winchester L, et al. (2023) Harnessing the potential of machine learning and artificial intelligence for dementia research. Brain Informatics 10. [crossref]
  12. Osaki NT, Ghosh S, Palaniappan SK, Maeda K (2024) Enabling personalization for digital cognitive stimulation to support communication with people with dementia: Pilot intervention study as a prelude to AI developm JMIR Formative Research 8. [crossref]
  13. Rasmussen HEGE, Hellzen OVE (2013) The meaning of long-term caregiving for patients with frontal lobe dementia. International Journal of Qualitative Studies on Health and Well-being 8. [crossref]
  14. Wylie MA, Shnall A, Onyike CU, Huey ED (2013) Management of frontotemporal dementia in mental health and multidisciplinary settings. International Review of Psychiatry 25: 230-236. [crossref]
  15. Likas A, Vlassis N, Verbeek JJ (2003) The global k-means clustering algorithm. Pattern recognition 36: 451-46. [crossref]
  16. Moskowitz HR (2012) ‘Mind Genomics’: The experimental, inductive science of the ordinary, and its application to aspects of food and feeding. Physiology & Behavior 107: 606-613. [crossref]
  17. Moskowitz HR, Gofman A, Beckley J, Ashman H (2006) Founding a new science: Mind Genomics. Journal of Sensory Studies 21: 266-307.
  18. Porretta S, Gere A, Radványi D, Moskowitz H (2019) Mind Genomics (Conjoint Analysis): The new concept research in the analysis of consumer behaviour and choice. Trends in Food Science & Technology 84: 29-33.
  19. Wu T, He S, Liu J, Sun S, Liu K, et al. (2023) A brief overview of ChatGPT: The history, status quo and potential future development. IEEE/CAA Journal of Automatica Sinica 10: 1122-1136.
  20. Rosness TA, Haugen PK, Engedal K (2008) Support to family carers of patients with frontotemporal dementia. Aging and Mental Health 12: 462-466. [crossref]

Update on the Management of Status Asthmaticus Gravidus and Acute Severe Asthma during Pregnancy

Leave a reply

DOI: 10.31038/AWHC.2024712

Introduction

As noted in our recent review on status asthmaticus gravidus [1], a quarter of pregnant women with asthma will experience acute severe exacerbations of resulting in emergency department visits or hospitalizations [2,3]. There is wide variability in asthma control during pregnancy [4]. Overall, approximately a third of pregnant patients experience worse asthma control, one third will have clinical improvement and one third will experience no change [5]. Importantly, nearly half of pregnant women experience acute asthma exacerbations requiring emergency care. Since publication of our overview on the management of acute severe asthma in pregnancy last year, there has been an update to the Global Initiative for Asthma (GINA) guidelines. In addition, national consensus guidelines have been published for women with asthma in China as well as for pregnant patients with asthma in Brazil. The purpose of this brief update is to highlight recent changes relevant for the changes for the management of acute severe asthma in pregnant patients, including new research findings and opportunities.

Acute Asthma Management in Pregnancy Update

All of these asthma guidelines emphasize the need for rapid and aggressive interventions to treat severe acute asthma exacerbations in pregnant women in order to minimize the risks of hypoxia to both the mother and fetus. Consistent with the updated GINA guidelines, the Chinese and Brazil consensus guidelines each note that drug therapy for acute asthma exacerbations in pregnant women is similar to that of nonpregnant women, including the use of inhaled beta2-agonists, inhaled ipratropium, and administration of systemic corticosteroids. Both national guidelines note that safety data are generally lacking in pregnant patients for “many drugs for treating asthma”. Furthermore, the Brazilian guidelines cite our article on status asthmaticus gravidarum, noting that this life-threatening asthma syndrome may require additional therapies, such as magnesium sulfate, that have “limited efficacy data in pregnant patients” [6-8].

COVID in Pregnancy Update

Pregnant patients with asthma have a higher incidence of severe respiratory viral infections. A recent report out of Denmark suggests that there is an increased risk of infection with SARS-Co-V-2 in pregnant patients with asthma compared to those without asthma. Furthermore, pregnant patients with asthma have a seven-fold increased risk of severe complications with SARS-Co-V-2 infection compared to pregnant patients without asthma. In contrast, patients with asthma do not have a higher risk of complications among non-pregnant patients hospitalized with SARS-Co-V-2. Indeed, reports have suggested that asthma may be protective against SARS-Co-V-2 infection due to a reduction in angiotensin-converting enzyme (ACE)-2 receptor expression and reduced viral entry due to Type 2 cytokines such as Interleukin (IL)-13. Understanding the reasons between pregnant and non-pregnant responses to SARS-CoV-2 is worthy of additional investigation. [9-13]

Research Needs

Importantly, there have been no further clinical trials published on asthma management strategies in acute severe asthma in pregnant patients. The GINA updates noted “the need for greater clarity in current recommendations and the need for more randomized clinical trials (RCTs) among pregnant asthma patients” [14]. Thus, there remains a need to further examine the role of additional pharmacologic agents, especially biologics, in the management of acute severe asthma in pregnancy. Importantly, the Brazil guidelines also mention the important role of phenotyping asthma to optimize disease management and treatment choice. Though the authors note that “identifying the primary phenotype as allergic or non-allergic may be enough”. As noted above, there is wide variation in the disease course of pregnant patients with asthma. Development of patient-specific phenotypes may identify pregnant asthmatic patients that would benefit from individualized acute treatment, specifically anti-inflammatory biologics.

Disclosure Statement

Dr. Cairns has no disclosures directly related to the topic of asthma. He has served as a consultant for bioMerieux for the development and use of biomarkers and he has received grant support from the National Institutes of Health (NIAID, NHLBI) and the Bill and Melinda Gates Foundation for COVID-19 studies and interventions.

Dr. Kraft has received funds paid directly to the institution for research in asthma by the National Institutes of Health, American Lung Association, Arteria, and Sanofi-Regeneron. She has served as a scientific consultant with funds paid to her to address pathobiology of asthma for AstraZeneca, Sanofi-Regeneron, Chiesi Pharmaceuticals, Kinaset and Genentech. Dr. Kraft is also co-founder and Chief Medical Officer for RaeSedo, Inc. created to develop peptidomimetics for the treatment of inflammatory lung disease. The company is currently in the pre-clinical phase of therapeutic development.

References

  1. Cairns CB, Kraft M (2023) Status Asthmaticus Gravidus: Emergency and Critical Care Management of Acute Severe Asthma During Pregnancy. Immunol Allergy Clin North Am 43: 87-102. [crossref]
  2. Hasegawa K, Craig SS, Teach SJ, Camargo CA (2021) Management of Asthma Exacerbations in the Emergency Department. J Allergy Clin Immunol Pract 9: 2599-610. [crossref]
  3. Enriquez R, Griffin MR, Carroll KN, Wu P, Cooper WO, et al. (2007) Effect of maternal asthma and asthma control on pregnancy and perinatal outcomes. J Allergy Clin Immunol 120: 625-30. [crossref]
  4. Labor S, Tir AMD, Plavec D, Juric I, Roglic M, et al. (2018) What is safe enough – asthma in pregnancy – a review of current literature and recommendations. Asthma Research and Practice 4: 11. [crossref].
  5. Kircher S, Schatz M, Long L (2002) Variables affecting asthma course during pregnancy. Ann Allergy Asthma Immunol 89: 463-466. [crossref]
  6. GINA Global Initiative for Asthma (GINA)(2023) Global Strategy for Asthma Management and Prevention
  7. Hu Q, Chen X, Fu W, Fu Y, He K, et al. (2024) Chinese expert consensus on the diagnosis, treatment, and management of asthma in women across life. J Thorac Dis 20 16: 773-797. [crossref]
  8. Carvalho-Pinto RM, Cançado JED, Caetano LSB, Machado AS, Blanco DC (2023) Asthma and pregnancy. Rev Assoc Med Bras 69(1): e2023S123. [crossref]
  9. Bonham CA, Patterson KC, Strek ME (2018) Asthma outcomes and management during pregnancy Chest 153: 515-27. [crossref]
  10. Aabakke AJM, Petersen TG, Wøjdemann K, Ibsen MH, Jonsdottir F, et al. (2023) Risk factors for and pregnancy outcomes after SARS-CoV-2 in pregnancy according to disease severity: A nationwide cohort study with validation of the SARS-CoV-2 diagnosis. Acta Obstet Gynecol Scand 102: 282-293. [crossref]
  11. Ozonoff A, Schaenman J, Jayavelu ND, Milliren CE, Calfee CS, et al. (2023) IMPACC study group members. Phenotypes of disease severity in a cohort of hospitalized COVID-19 patients: Results from the IMPACC study. EBioMedicine 83: 104208.
  12. Kimura H, Francisco D, Conway M, Martinez FD, Vercelli D, et al. (2020) Type 2 inflammation modulates ACE2 and TMPRSS2 in airway epithelial cells. J Allergy Clin Immunol 146: 80-88.e8. [crossref]
  13. McPhee C, Yevdokimova K, Rogers L, Kraft M (2023) The SARS-CoV-2 pandemic and asthma: What we have learned and what is still unknown. J Allergy Clin Immunol 152: 1376-1381. [crossref]
  14. McLaughlin K, Foureur M, Jensen ME, et al. (2018) Review and appraisal of guidelines for the management of asthma during pregnancy. Women Birth 31: e349-e357. [crossref]
fig 1

Graphite and Diamond-Rich Pegmatite as a Small Vein in a Gneiss Drill Core from the Annaberg Region/ Erzgebirge, Germany

Leave a reply

DOI: 10.31038/GEMS.2024622

Abstract

Diamond and graphite in a vertical pegmatite veinlet in a gneiss drill core from the Annaberg region/Erzgebirge, Germany, demonstrate a more crustal position and underline the greater meaning of the input of supercritical fluids from mantle deeps. Proof for that statement is a high concentration of nano-diamond-bearing graphite as a micrometer to sub-micrometer large crystals in quartz and orthoclase.

Keywords

Pegmatite, Nanodiamond, Graphite, Raman spectroscopy, Supercritical fluids

Introduction

End of the eighties, we studied drill cores, most granites from the Annaberg district, for melt inclusion to reconstruct the temperature and pressure of granites with cassiterite-bearing vein/veinlet structures. Most samples are from the borehole An 10/85 near the Grundteichschenke north of Schlettau (Buchholz region). Some results are used and cited in Hösel et al. (1992) [1]. The primary data are in the Thomas (1988) [2]. Most samples are granite drill cores. The solidus, liquidus, homogenization temperatures, and water content of melt inclusions were determined on these granite samples in 1988. However, sample T6 from the drill core An 10/85 (at 71.0 m depth) is a gneiss core with a 2 cm thick pegmatite veinlet parallel to the drill core axis (perpendicular). The pegmatite veinlet was by the field geologist as a dolomite-fluorite vein wrongly interpreted. A microscopic study showed that dolomite and fluorite are not present. Only quartz, feldspars, muscovite, zircon, and apatite are observable. The sample was not studied then because of the complete absence of fluid and melt inclusions.

Methods

We used for all microscopic and Raman spectrometric studies a petrographic polarization microscope with a rotating stage coupled with the RamMics R532 Raman spectrometer working in the spectral range of 0-4000 cm-1 using a 50 mW single mode 532nm laser. Details are in Thomas et al. 2022a and 2022b [3,4]. We used the Olympus long-distance LMPLN100x for the Raman spectroscopic routine measurements as a 100x objective. To avoid contamination on the sample surface, we studied only mineral grains deep under the surface. Therefore, we generally used the total laser power of 50 mW on the sample and a long counting time of 100 to 200 seconds and sometimes up to 10 minutes.

Sample

Figure 1 shows the used sample T6 from the drill core An 10/85 (at 71.0 m). Besides the typical pegmatite minerals quartz, feldspars, mica, zircon, zircon-reidite, xenotime-(Y), monazite-(Ce), and apatite, the pegmatite is characterized by an unexpectedly large amount of graphite and nanodiamond grains in quartz (Figure 2) and also in orthoclase. Fluid inclusions (≤ 2 µm) are very rare. The 500 µm thick sample is polished on both sides with an Al2O3-H2O suspension and carefully cleaned. Graphite and diamond are the objects of this short study. To avoid possible contamination by diamonds from the drilling and preparation, we used only graphite and diamonds deeper than 30 µm (see Thomas et al. 2023).

fig 1

Figure 1: View of the drill core sample. a) Side view of the gneiss core with yellow-brown pegmatite veinlet. X shows the sample position. b) Top view of the sample. The red and black crystals are hematite and pyrrhotine, respectively.

fig 2

Figure 2: Distribution of the graphite grains in pegmatite quartz about 30 µm deep from the sample surface.

Results

During the microscopic study of the quartz and orthoclase from the pegmatite, black crystals and aggregates (≥ 5 µm) are noticeable. The mean is 1.8 x 108 black grains per cm3 in quartz. The distinction between graphite and nanodiamond is only possible with the Raman spectroscopy. There are sporadic also graphite-diamond grains present in feldspar. Figure 2 shows the distribution of graphite-like crystals and clusters in quartz.

The distribution is not homogeneous. Partial hydrothermal re-mobilised quartz (lighter) is pure in graphite. However, the 1580 cm-1 Raman graphite band is also present at long counting times in transparent quartz regions. That means the sample has a very high density of tiny to nano-graphite particles-invisible at the highest magnification (100x ocular). Table 1 gives the Raman data for “invisible” diamonds and graphite in transparent quartz regions, and Figure 3 shows the corresponding Raman spectrum.

Table 1: Raman band of diamond and graphite in clear pegmatite quartz of the sample

Mineral phase

 Raman band (cm-1)

FWHM (cm-1)
Quartz

1233.2

17.4
Diamond

1333.1

19.1
Graphite D1

1353.6

33.2
Graphite G

1580.7

28.0
Graphite D2

1615.7

37.1

FWHM: Full Width at Half Maximum

fig 3

Figure 3: Raman spectrum of clear pegmatite quartz without microscopically visible graphite. Qtz-weak quartz band at 1233.2 cm-1. Recording conditions: 50 mW on sample, exposition time of 10 minutes, 100x objective (see Table 1).

Under the more macroscopical black dots (often spherical or elliptical), there are primarily mixtures of diamond and graphite-Figure 4 shows such crystals. The large xenotime-(Y) crystal in Figure 4a is conspicuous. The Raman spectra match 97% of the xenotime-(Y) RRUFF database ID: R050178 [5]. Besides xenotime-(Y), there are also monazite-(Ce) crystals present [RRUFF database ID: R040106, match 95%], mostly in larger graphite aggregates. Both REE minerals are primarily present in larger graphite-nanodiamond crystal clusters, demonstrating that these minerals are also related to the fast-rising supercritical fluids. Table 2 shows the obtained Raman data (Gaussian fit) of the studied graphite and nanodiamond in Figures 4 and 5.

Table 2: Raman data for the graphite-diamond aggregate shown in Figure 4a

Mineral phase

 Raman band (cm-1)

FWHM (cm-1)
Diamond tip1)

1267.8

71.6
Diamond

1322.6

46.9
Diamond (bulk)

1333.1

100.0
Graphite D1

1352.7

56.8
Graphite G

1571.8

81.9
Graphite D2

1615.7

39.3

FWHM: Full Width at Half Maximum. 1)According to Zaitsev (2001), this range is typical for isolated crystallites of diamonds – here, nanodiamonds.

fig 4

Figure 4: Graphite-nano-diamond aggregates in pegmatite quartz from the drill core T6 from the drilling An 10/85. a) The crystal with Xtm marked place in the graphite-nano-diamond aggregate is a xenotime-(Y) crystal. The crystal shown in b) is composed only of graphite and diamond.

fig 5

Figure 5: Raman spectrum of a graphite-diamond aggregate (Figure 4a) in pegmatite quartz (sample T6). Shown are only the principal data. More information is in Table 2.

Table 3 summarizes the Raman data of nanodiamond and graphite in the pegmatite sample T6.

For quartz, the Raman band ranges from 1327.3 to 1351.1 cm-1. According to Zaitsev (2001) [6], this range is typical for isolated crystallites of diamonds-here, nanodiamonds with grain size in the range of several nanometers. Besides the spheric to elliptic graphite aggregates, there are also whisker-like graphite needles (see Figure 6); however, there are never moissanite whiskers.

Table 3: Raman data for nanodiamond (n=20) and graphite (n=10) in quartz and orthoclase in the pegmatite from samples T6, 20, and 10 different crystals, respectively.

 

 Diamond Graphite
Mineral Raman band (cm-1) FWHM (cm-1) Raman band (cm-1)

FWHM (cm-1)
Quartz

1339.4 ± 12.1

 41.8 ± 12.0 1580.3 ± 4.5

1615.4 ± 3.8

39.2 ± 11.9

29.2 ± 4.3
Orthoclase

1337.5 ± 6.8

 49.9 ± 16.3

1571.8 ± 8.0

27.8 ± 4.0
Or-matrix*

1352.4

 31.5 1581.7 27.9

*Free of visible graphite (50mW on sample, 10 minutes recording time)

fig 6

Figure 6: Graphite needles or whiskers beside graphite-bearing nanodiamond cluster. Gr: Graphite, nD: Nanodiamond. Note the needles are real needles and not sections of flat graphite crystals.

Discussion

In the last couple of years, the author, with his colleagues, found in different Variscan granites, pegmatites, and related mineralizations from more crustal position minerals like diamond, graphite, moissanite, reidite, coesite, stishovite, and others, representing mantle origin. By the dominance of spherical forms and their extraneous position in the host minerals, fast transport via supercritical fluids is almost imperative. Proofs are in Thomas et al. (2023a) [7] and Thomas (2023a and 2023b) [8,9]. The small vertical pegmatite vein in gneiss (sample T6) with graphite and nanodiamond further hints that supercritical fluids play a more significant role than assumed. The search for moissanite (isometric crystals or whiskers) in the quartz of the given sample was unsuccessful. Therefore, we can conclude that the formation of moissanite whiskers and isometric crystals in the beryl-dominant veins of the Sauberg mine near Ehrenfriedersdorf beryllium and water are essential catalysts for the formation of moissanite [10]. The diamond (nanodiamond) and graphite spectra look like the shock-synthesized diamond by Chen et al. (2004), representing strongly nonequilibrium processes during the change of the supercritical state into a critical/undercritical one. Most diamonds/nanodiamonds show a covering by graphite. All spectra differ from static pressure diamonds [11]. The relatively extended stay at high temperatures makes the primary diamond unstable and transforms it into nanodiamond or onion-like carbon (OLC) – see Zou et al. 2010 [12]. The fine-disperse distribution of nanographite and nanodiamond in the quartz and orthoclase matrix is conspicuous. Maybe these prevent the intense formation of fluid inclusions in quartz and orthoclase during the cooling.

Acknowledgment

Günter Hösel (Freiberg) is thanked for providing the drilling sample material from the Annaberg

References

  1. Hösel G, Kühne R, Zernke B (1992) Zur Zonalität der Zinnmineralisation im Raum Annaberg/Erzgebirge. Geoprofil 4: 49-57.
  2. Thomas R (1988) Ergebnisse der thermobarometrischen Untersuchungen an Granitproben aus dem Gebiet Annaberg. Unpublished Report, Freiberg.
  3. Thomas R, Davidson P, Rericha A, Recknagel U (2022a) Discovery of stishovite in the prismatine-bearing granulite from Waldheim, Germany: A possible role of supercritical fluids of ultrahigh-pressure origin. Geosciences 12: 1-13.
  4. Thomas R, Davidson P, Rericha A, Voznyak DK (2022b) Water-rich melt inclusions as “frozen” samples of the supercritical state in granites and pegmatites reveal extreme element enrichment resulting under nonequilibrium conditions. Min J (Ukraine) 4 4: 3-15.
  5. Lafuente B, Downs RT, Yang H, Stone N (2016) The power of database. The RRUFF project. In: Armbruster T, Danisi RM (Eds.), Highlights in Mineralogical Crystallography. De Gruyter, Berlin, München, Boston, USA, Pg: 1-30.
  6. Zaitsev AM (2001) Optical Properties of Diamond. A data Handbook. Springer-Verlag Berlin Heidelberg GmbH 502.
  7. Thomas R (2023a) Ultrahigh-pressure and-temperature mineral inclusions in more crustal mineralizations: The role of supercritical fluids. Geol Earth Mar Sci 5: 1-2.
  8. Thomas R, Davidson P, Rericha A, Recknagel U (2023a) Ultrahigh-pressure mineral inclusions in a crustal granite: Evidence for a novel transcrustal transport mechanism. Geosiences 13: 1-13.
  9. Thomas R, Recknagel U, Rericha A (2023b) A moissanite-diamond-graphite paragenesis in a small beryl-quartz vein related to the Variscan tin-mineralization of the Ehrenfriedersdorf deposit, Germany. Aspects Min Miner Sci 11: 1310-1319.
  10. Thomas R (2023b) The Königshainer granite: Diamond inclusion in zircon. Geol Earth Mar Sci 5: 1-4.
  11. Chen P, Huang F, Yun S (2004) Structural analysis of dynamically synthesized diamonds. Materials Research Bull 39: 1589-1597.
  12. Zou Q, Wang MZ, Li YG, Lv B, Zhao YC (2010) HRTEM and Raman characterization of the onion-like carbon synthesised by annealing detonation nanodiamond at lower temperature and vacuum. J Experim Nanosci 5: 473-487.
fig 1

REE-rich Fluorite in Granite from Zinnwald/East Erzgebirge/Germany

Leave a reply

DOI: 10.31038/GEMS.2024621

Abstract

The REE-rich fluorites in quartz of the topaz-albite granite from Zinnwald/Erzgebirge are often related to nanodiamonds and graphite. Together, the solvus curves (water content of melt inclusions in granite quartz versus temperature) and the Lorentzian element distribution (F, Rb, Cs) prove the input of supercritical fluids and their influences on the element redistribution in the granite. Together with the impact of supercritical fluids, the crystallization history of the topaz-albite granite from Zinnwald is very complex.

Keywords

Topas-albite-granite, REE-rich fluorites, Fluocerite and tveitite trends, Nanodiamonds, Graphite, Raman spectroscopy

Introduction

During the study of melt and fluid inclusions in quartz and topaz from the Zinnwald granite [1] we often found in granite quartz spherical crystals of REE-rich fluorites beside other, for “normal” granites untypical mineral phases: magmatic fluorite, cryolite, elpasolite, and rubidian leucite with the empiric formula (K0.64Rb0.22,Na0.13Cs0.01)(Al0.96Fe0.03)Si2O6, and boromuscovite. According to melt inclusion results, the fluorine concentration in the melt of the evolved granite phases increases to 5.64 ± 0.19%(g/g). It was also essential that for this granite, we could construct from the analytically determined water concentration of different melt inclusion a pseudobinary XH2O vs. T plot of re-homogenized type-A and type B-melt inclusion with a solvus crest at 720°C and 28.6%(g/g) H2O (Figure 10 in there). That was a natural granite system’s first pseudobinary solvus curve [1]. In the meantime, we have seen that such pseudobinary solvus curves are mostly connected with the extreme enrichment of some elements. For the case of Zinnwald, we found Lorentzian curves for F, Rb, and Cs [2]. Such curves are strong proof of the participation of supercritical fluids. However, around 2005, nobody had any idea about the role of supercritical fluids in granite formation and mineralization. Using the small example of REE-rich fluorite globules in quartz, we will show that supercritical fluids play an essential part in granite formation and re-crystallization.

Methods

Primary for the first identification of the REE-rich fluorites, we used a Dilor XY Laser Raman Triple 800 mm spectrometer equipped with an Olympus optical microscope. The spectra were collected with a Peltier-cooled CCD detector using a laser wavelength of 488 nm. For recent studies, we used for all microscopic and Raman spectrometric studies a petrographic polarization microscope with a rotating stage coupled with the RamMics R532 Raman spectrometer working in the spectral range of 0-4000 cm-1 using a 50 mW single mode 532nm laser. Details are in Thomas et al., 2022a and 2022b [2]. For the Raman spectroscopic routine measurements, we used the Olympus long-distance LMPLN100x as a 100x objective. We carefully cleaned the samples to prevent diamond contamination due to the preparation. For the Raman determination, we used only 30 or more µm deep crystals from the sample surface [3]. One nanodiamond sample (Figure 5) is on the surface. However, the Raman lines are characteristic of nanodiamonds, not contaminated diamonds [3]. To determine the composition of the minerals in question, we used the microprobes CAMECA SX 50 and SX100. Details are in Franz et al. (1996) [4-7]

Sample

The samples (TH212) are on both sides, about 500 µm thick polished granite sections (Figure 1). The used sample is a topaz-albite granite collected as a boulder about 1.3 km from Fuchshübel, about 1.3 km northeast of Zinnwald in the East-Erzgebirge. A concise description was given by Thomas et al. (2005) [1]. The quartz contains tiny, mostly spherical colorless crystals of REE-rich fluorite. Other typical and untypical minerals are graphite-whiskers, F-rich needle-like topaz crystals, and orthorhombic cassiterite with graphite and nanodiamond inclusions. Thomas (2023) [8], as well as fluorite, cryolite, elpasolite, rubidian leucite, and boromuscovite (all in quartz). For comparison, we used an emerald-green REE-bearing fluorite from the Sachsenhöhe near Zinnwald [9]. Note that the very smooth surface of the small diamond spheres at the sample surface causes these grains to fall out easily during preparation (polishing).

Another relatively REE-rich fluorite (No. Z 9054) was an emerald-green piece as big as your fist from the Sachsenhöhe near Zinnwald [9]. This sample served to calibrate the ICP-AES and microprobe SX50 instruments of the GFZ in the nineties. The sample contains 0.27% Y and 0.30% REE. Another REE-rich fluorite is from Ehrenfriedersdorf sample Sn70 with a maximum of 0.90% Y and 1.08% REE.

fig 1

Figure 1: Thick section ZW-TH212-I (500 µm thick and on both sides polished). Scale is in centimeters. The colorless parts are quartz with REE-rich fluorite globules.

Results

The REE-rich fluorite globules in the Zinnwald granite quartz have diameters of around 20 µm (up to 70 mm) and are only present in the larger quartz crystals of the granite (Figure 1). Figure 2 shows the form of the typical REE-rich fluorites and Figure 3 a typical Raman spectrum.

fig 2

Figure 2: Typical REE-rich spherical fluorite crystals in Zinnwald quartz. The scale (below right) is valid for all examples in the Figure. Gr: Graphite and nanodiamonds in graphite.

fig 3

Figure 3: Typical Raman spectrum of REE-rich fluorite in granite quartz from Zinnwald. Note that the characteristic strong Raman band at 321 cm-1 for REE-free fluorite is missing or has been shifted to 242 cm-1.

The interpretation of REE-rich fluorite Raman spectra is difficult. There are too many variables (Y vs. sum of REE. Only two passable correlations exist between the sum of REE (in at%) and the Raman band position at about 650 cm-1 and the fluorine content, respectively. The position of the second broad Raman band (between 400 and 550 cm-1) correlated roughly with the Y concentration. More work is, however, necessary, primarily because this Raman band is a double band. Often, these fluorite globules contain remnants of nanodiamond-bearing graphite inside, or the graphite forms small rims around the crystals. Besides the spherical REE-rich fluorites, there are also graphite globules (Figure 4a) up to 40 µm in diameter. Smaller graphite spheres with remnants of diamonds are rarer (Figure 4b) to see. The spherical form of the REE-rich fluorites in quartz, nanodiamond, and graphite indicate clearly that these crystals are unambiguous foreign crystals in the granite quartz.

fig 4

Figure 4: Spherical graphite crystal a) with nanodiamonds in Zinnwald quartz. b) graphitized diamond in the same quartz (about 30 µm deep). The Raman band of the diamond is 1320.2 cm-1 with a FWHM of 56.7 cm-1. FWHM: Full Width at Half Maximum.

Besides the REE-rich fluorite globules, fluorite with larger graphite aggregates is also present (Figure 5). Also, this graphite aggregate contains nanodiamonds. The presence of fluorite-cryolite-topaz aggregates (see Thomas et al., 2005 [1]; Figure 1c in it) in the quartz of this peraluminous granite rock clearly shows a second peralkaline history. Additionally, the nanodiamonds, combined with the solvus and some Lorentzian-distributed elements [10], prove the input of supercritical fluids from the deeper mantle region. In this short contribution, we will only concentrate on the REE-rich fluorites. Table 1 shows the results of the microprobe analyses on the REE-rich fluorites. The abbreviation of the REEs in Table 1 is Ln (the sum of the REE). It is a selection of data obtained in more than 15 years. They show, however, the characteristic properties of these fluorites. Primarily, the author thinks the REE-rich fluorites are a tveitite-like mineral. However, tveitite-(Y) [symplified as Ca14Y5F43] is hexagonal. The lattice parameters using the TEM technique (unpublished data by R. Wirth, GFZ Potsdam and Wirth, 2004) [11] determined for a, b, and c identical values of 5.46 Å, also a cubic mineral. The density is 4.0 g/cm3. Also, birefringence is missing. In contrast to simple fluorites, the REE-rich ones show in no case the typical 321 cm-1 Raman line for pure fluorites. Figure 6 shows a REE-rich fluorite crystal like a twinning. On the right side, there are tiny, very REE-rich microcrystals (no monazite or xenotime!). The Raman spectrum is similar; however, there are also more significant Raman band differences.

fig 5

Figure 5: Fluorite (Fl) with graphite in Zinnwald granite quartz. The Raman band for fluorite is at 321.7 cm-1, for diamond at 1327.2 cm-1, and for graphite at 1345.3 cm-1 (D1), 1576.2 cm-1 (G), and 1602.4 cm-1 (D2) with the FWHM values of 9.7, 44.9, 71.2, 73.7, 38.8, respectively.

Table 1: Chemical composition (in % (g/g) of REE-rich fluorites from Zinnwald (reduced number)

Component

 ZW-1 ZW-2 ZW-3 ZW-4 ZW-5 ZW-6 ZW-7 ZW-8 ZW-9 ZW-10

ZW-11
Na

0.24

 0.23 0.23 0.23 0.31 0.24 0.12 0.00 0.00 0.16 0.14

Ca

 36.99 36.75 36.77 36.97 35.45 36.86 32.75 45.97 39.21 26.97

22.26
Y

4.43

 4.47 4.44 4.48 4.84 4.57 2.43 5.28 0.86 3.09 2.81
La

1.59

 1.65 1.65 1.65 1.78 1.67 5.10 1.79 5.12 5.99

5.89
Ce

4.67

 4.76 4.63 4.64 5.13 4.60 13.83 5.06 13.45 16.20 15.97

Pr

 0.62 0.66 0.60 0.55 0.87 0.58 1.50 0.65 1.65 1.76

1.74
Nd

1.75

 2.03 2.08 2.19 2.56 2.10 4.92 2.26 4.35 5.74 5.69

Sm

 0.57 0.58 0.60 0.56 0.61 0.55 1.05 0.61 1.45 1.22

1.22
Eu

0.00

 0.00 0.00 0.00 0.00 0.00 0.00 0.01 0.00 0.00

0.00
Gd

0.91

 0.94 0.97 0.91 0.92 0.90 0.77 0.99 0.53 0.89 0.89

Tb

0.14

 0.12 0.13 0.16 0.00 0.10 0.17 0.14 0.02 0.20 0.00

Dy

0.84

 0.88 0.92 0.91 1.03 0.91 0.60 0.95 0.34 0.68 0.69
Ho

0.24

 0.15 0.22 0.19 0.00 0.19 0.05 0.21 0.05 0.05 0.05

Er

0.66

 0.84 0.75 0.84 0.00 0.71 0.25 0.81 0.16 0.28 0.29

Tm

0.00

 0.00 0.00 0.00 0.00 0.00 0.00 0.13 0.00 0.00 0.00

Yb

0.94

 0.93 0.95 0.93 0.00 0.96 0.47 1.00 0.07 0.54 0.55

Lu

0.13

 0.15 0.16 0.15 0.00 0.14 0.10 0.16 0.02 0.12 0.12

F

45.66

 44.28 44.90 44.66 46.49 44.93 36.12 33.99 32.72 36.12 41.72

Total

100.4

 99.42 100.0 100.0 99.99 100.0 100.2 100.0 100.0 100.01

100.0
Formula coefficients calculated based on one cation
Na

0.010

 0.010 0.010 0.010 0.140 0.010 0.005 0.000 0.000 0.010 0.006

Ca

0.919

 0.917 0.917 0.922 0.885 0.920 0.814 1.147 0.978 0.67 0.56

Y

0.050

 0.500 0.050 0.050 0.054 0.051 0.027 0.059 0.010 0.03 0.032
La

0.011

 0.012 0.012 0.012 0.013 0.012 0.037 0.013 0.037 0.04 0.042
Ce

0.033

 0.034 0.033 0.033 0.037 0.033 0.098 0.036 0.096 0.12 0.114

Pr

0.004

 0.005 0.004 0.004 0.006 0.004 0.011 0.005 0.012 0.01 0.012

Nd

0.012

 0.014 0.014 0.015 0.018 0.015 0.034 0,016 0.030 0.04 0.039

Sm

0.004

 0.004 0.004 0.004 0.004 0.004 0.007 0.004 0.010 0.01 0.008

Eu

0.000

 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.00 0.000

Gd

0.006

 0.006 0.006 0.006 0.006 0.006 0.005 0.006 0.003 0.01 0.006

Tb

0.001

 0.001 0.001 0.001 0.000 0.001 0.001 0.001 0.000 0.00 0.000
Dy

0.005

 0.005 0.006 0.006 0.006 0.006 0.004 0.006 0.002 0.00 0.004

Ho

0.001

 0.001 0.001 0.001 0.000 0.001 0.000 0.001 0.000 0.00 0.000

Er

0.004

 0.005 0.004 0.005 0.000 0.004 0.001 0.005 0.001 0.00 0.002

Tm

0.000

 0.000 0.000 0.000 0.000 0.000 0.000 0.001 0.000 0.00 0.000
Yb

0.005

 0.005 0.005 0.005 0.000 0.006 0.003 0.006 0.000 0.00 0.003

Lu

0.001

 0.001 0.001 0.001 0.000 0.001 0.001 0.001 0.000 0.00 0.001

F

2.394

 2.331 2.363 2.351 2.447 2.365 1.894 1.789 1.722 1.90 2.196

SLn

0.087

 0.093 0.09 0.090 0.090 0.090 0.202 0.191 0.191 0.23 0.23

S(Y+Ln)

0.137

 0.098 0.14 0.140 0.144 0.140 0.229 0.201 0.201 0.26

0.26

Ln: Measured Lanthanides

fig 6

Figure 6: REE-rich fluorite – a) optical microphotography and b) electron microprobe BSE image. D – diamond and graphite. The left lower part in b) shows the immiscibility of heavy-REE-rich fluorites. In Figure 6a, black points in the right part of the crystal are not graphite, which is seen in the BSE image (Figure 6b).

The nanodiamond (D) at the left upper corner has the main band at 1332.8 cm-1 with a FWHM = 47.9 cm-1, and the graphite bands are positioned at 1363.1, 1571.4 1602.3 cm-1 with the FWHM of 40.5, 44.0, and 40.2 cm-1 respectively.

Figure 7 shows Y + Ln versus Ln (values in at%) with the zero point for pure water-clear synthetic fluorite (not shown in the table). The circles are the data from Table 1 plus unpublished data [1], and the triangles represent the data for REE-rich fluorites and tveitite-(Y) from Pekov et al., 2009 [12]. There are two trends: the dashed blue line [12] corresponds to the tveitite-(Y) trend, and the dashed black line corresponds to the fluocerite trend for the REE-rich Zinnwald fluorites.

fig 7

Figure 7: (Y + Ln) versus Ln (all in at%) of REE-rich fluorites. Ln = sum of the REE, (Y + Ln) = sum of Y + REE’s. The black circles (Zinnwald) and the dashed black line correspond to the fluocerite trend (this work), and the blue triangles and dashed blue line equal the tveitite trend, according to Pekov et al. (2009) [12].

The simultaneous crystallization of xenotime-(Y) and monazite-(Y) together with the REE-rich fluorite prevents the crystallization of the fluocerite [12] standing at the end of the trend (fluocerite trend of Figure 7). That is also true for other REE-rich fluorite in the region and around Ehrenfriedersdorf. The emerald-green fluorite (Z 9054) from the Sachsenhöhe near Zinnwald contains 0.29% (g/g) Y and 0.31% (g/g) REE, and the fluorite from Ehrenfriedersdorf (sample Sn70) has 0.60% (g/g) Y and 1.22% (g/g) REE’s. Compared with this, the REE-rich Zinnwald fluorite contains up to 5.3% (g/g) Y and 33.1% (g/g) REE’s. The high REE content of the fluorite here is the result of the supercritical fluids in the Zinnwald region being deleted in phosphorus [12]. Figure 8 shows the chondrite standardized REE distribution of the two different fluorites from Zinnwald (ZW) and the Sachsenhöhe near Zinnwald [13].

By the simultaneous crystallization of more significant amounts of monazite-(Y) and xenotime-(Y) in the case of the fluorite from the Sachsenhöhe, is the concentration of all REE significantly lower than in the REE-rich fluorites from Zinnwald.

fig 8

Figure 8: Chontrite standardized REE concentration in fluorites from Zinnwald (ZW) and the Sachsenhöhe (SH). The REE concentration in fluorite is in ppm. Y is at the position between dysprosium and holmium inserted there as the result of the stereochemistry behavior of yttrium [14].

Discussion

The topaz-zinnwaldite-albite granite from Zinnwald has obviously a very complex history [1]. Besides melt inclusions representing the crystallization of the topaz-zinnwaldite-albite granite, extreme water-rich melt inclusions are present (Table 3 in Thomas et al., 2005) [1], showing a more pegmatite-like state. At this time, the author also found REE-rich fluorite globules in quartz, for which no explanation could given. Later, we also found Lorentzian distributed elements (F, Rb, Cs) [10]. Now, during this study, the proven mantel indications (nanodiamond and graphite globules) tell us a further story – input of supercritical fluids coming from mantle deeps. These supercritical fluids are obviously very fluorine-rich, forming the REE-rich fluorite, and for a peraluminous granite, untypical minerals like cryolite, elpasolite, and rubidian leucite. The REE distribution patterns (Figure 8, ZW) differ clearly from the typical REE distribution patterns of hydrothermal and remobilized fluorites [14-16]. In a more hydrothermal later state, the minerals cryolite, cryolithionite, elpasolite, and native sulfur are daughter minerals in fluid inclusions in hydrothermal quartz crystals. Up to now, we could prove that supercritical fluids are active in the whole Erzgebirge, Slavkovsky les, the Saxon Granulite, Lusatian Granodiorites and quartz veins, Königshain Granite Massifs.

Acknowledgment

Thanks to Dr. V. Grunewald (ZGI Berlin) for the fluorite sample Z 9054. For microprobe analyses, we thank D. Rhede, H.-J. Förster, and Ona Appelt (all GFZ) for their help in the very past.

References

  1. Thomas R, Förster HJ, Rickers K, Webster JD (2005) Formation of extremely F-rich hydrous melt fractions and hydrothermal fluids during differentiation of highly evolved tin-granite magmas: a melt/fluid-inclusion study. Contrib Mineral Petrol 148: 582-601.
  2. Thomas R, Davidson P, Rericha A, Voznyak DK (2022) Water-rich melt inclusions as “frozen” samples of the supercritical state in granites and pegmatites reveal extreme element enrichment resulting under non-equilibrium conditions. Min J(Ukraine) 44: 3-15.
  3. Thomas R, Davidson P, Rericha A, Recknagel U (2023) Ultrahigh-pressure mineral inclusions in a crustal granite: Evidence for a novel transcrustal transport mechanism. Geosciences 94: 1-13.
  4. Franz G, Andrehs G, Rhede D (1996) Crystal chemistry of monazite and xenotime from Saxothuringian-Moldanubian metapelites, NE Bavaria, Germany. Eur J mineral 8: 1097-1118.
  5. Seifert W, Thomas R, Rhede D, Förster HJ (2010) Origin of coexisting wüstite, Mg-Fe and REE phosphate minerals in graphite-bearing fluorapatite from the Rumburk granite. Eur J Mineral 22: 495-507.
  6. Breiter K, Förster HJ (2021) Compositional variability of monazite.cheralite-huttonite solid solution, Xenotime, and Uraninite in geochemically distinct granites with special emphasis to the strongly fractionated peraluminous Li-F-P-rich Podlesi granite system (Erzgebirge/Krušné Hory Mts., Central Europe. Minerals 127: 1-21.
  7. Thomas R, Davidson P, Rhede D, Leh M (2009) The miarolitic pegmatites from the Königshain: a contribution to understanding the genesis of pegmatites. Contribution to Mineralogy and Petrology 139: 394-401.
  8. Thomas R (2023) Unusual cassiterite mineralization, related to the Variscan tin-mineralization of the Ehrenfriedersdorf deposit, Germany. Aspects in Mining & Mineral Science 11: 1233-1236.
  9. Thomas R (1978) Thermobarometrische und kryometrische Untersuchungen an Fluorit- und Quarzproben aus einem Brekzienkörper der Sachsenhöhe bei Bärenstein/Osterzgebirge. Unpublished report for the ZGI Berlin, HV 623/75,1-14.
  10. Thomas R, Davidson P, Appel K (2019) The enhanced element enrichment in the supercritical states of granite-pegmatite systems. Acta Geochim 38: 335-349.
  11. Wirth R (2004) Focused ion beam (FIB): a novel technology for advance applications ofmicro- and nanoanalysis in geosciences and applied mineralogy. Eur J Mineral 16: 863-876.
  12. Pekov IV, Chukanov NV, Kononkova NN, Yakubovich OV, Massa W, et al. (2009) Tveitite-(Y) and REE-enriched fluorite from amazonite pegmatites of the western Keivy, Kola Peninsula, Russia: Genetic crystal chemistry of natural Ca, REE-fluorides. Geology of Ore Deposits 51: 595-607.
  13. Thomas R (1994) Fluid evolution in relation to the emplacement of the Variscan granites in the Erzgebirge region: A review of the melt and fluid inclusion evidence. Metallogeny of collisional Orogens. International Association of the Genesis of Ore Deposits (IAGOD), eds: By Seltmann, R., Kämpf, H., Möller, P., 70-81
  14. Bommer H (1941) Über die Einordnung des Yttriums in die Reihe der Lanthaniden. Zeitschrift für anorganische und allgemeine Chemie 248: 397-401.
  15. Möller P (1989) REE(Y), Nb, and Ta enrichment in pegmatites and carbonatite-alkalic rock complexes. In: Lanthanides, Tantalum and Niobium. Eds: by P. Möller, P. Černy, and Saupé, 380: 103-144
  16. Thomas R, Davidson P (2016) Origin of miarolitic pegmatites in the Königshain granite/Lusatia. Lithos 260: 225-241 and ESM.