Sociodynamics of Artificial People

Reads: 62  | Likes: 3  | Shelves: 0  | Comments: 4

  • Facebook
  • Twitter
  • Reddit
  • Pinterest
  • Invite

Status: Finished  |  Genre: Science Fiction  |  House: Booksie Classic

This is the technical report which sets the scene for my recent story, 'La Double Inconstance'. Translator picture.

Note

I used the research paper below (which I translated) as background to my recent story, La Double Inconstance. The text is quite technical and, at 3,000 words, quite long. AC.

---

arXiv.org > sociobio > arXiv:2107.05827
Sociobiology > Modelling (qualitative)
[Submitted on 13 Apr 2032]

Sociodynamics of Artificial People

Mireille Fossey (Université de Paris, Faculté des Sciences du vivant)

ABSTRACT

The high-profile emergence of artificial people (also known as models) into elite society has led to discussion as to implications and consequences. These debates have mostly been conducted with the social science paradigm of Critical Theory [1], defining tropes of specific and intersectional oppression. 

In this short overview we introduce a different framework of analysis, specifically the interdisciplinary fusion of population genetics, evolutionary psychology and marxian historiography. This paradigm provides a robust and well-grounded perspective both for the scientific understanding of the phenomena and for the generation of robust public policy prescriptions. Some conclusions are speculative. 

Submission history
From: Mireille Fossey [view email]
[v1] Tue, 13 Apr 2032 21:05:34 UTC (135 KB)

[Translated by Adam Carlton]

--- FULL TEXT --- 

Sociodynamics of Artificial People

Introduction

The recent prominence of artificial people, most particularly in the unsuccessful Mars mission, has led to considerable discussion and controversy. Artificial people, commonly known as 'models', have achieved leadership positions in the economy, in politics, the media, entertainment and the professions with consequential social tensions. 

To date most analysts have appealed to the framework of Critical Theory [1] formulating responses as those of oppression and intersectionality together with appeals to post-Enlightenment values to secure equitable outcomes. 

In contrast to this normative approach, in this paper we explore a paradigm rooted in evolutionary biology/psychology and marxian historiography which we believe provides a scientifically well-founded and grounded analysis. Benefits of our approach include better understanding of observed social dynamics and a more profound set of public policy prescriptions. 

Background

The arrival of artificial people - models - in elite positions has posed a challenge to baseline humans - hereafter termed baselines. Should baselines welcome these new additions or under some circumstances could they be considered a threat?

Critical Theory frames this as a problem of values or morality. Some absolute moral framework is sought to make intuitively-desired outcomes compelling. But in a universe conformant to the rules of physics there can be no absolute morality: sociobiology tells us that questions of morality and values in fact serve to codify the interests of specific  groups.

An example

We start with a simple problem. Consider an isolated island on which a population of seabirds dwells and reproduces without serious predators. One day a population of rats is introduced: what will be the results?

This has happened many times. The most likely outcome is that the rats will hunt the seabirds to extinction by predating their young. If the birds fight the rats and succeed in eliminating them, then the population will survive. If they don’t fight, or fight and fail, then they will not.

These variant outcomes are not the concern of physics: probably the result is opaquely predetermined by the initial conditions. But genes ‘care’: the evolutionary process selects for genes/alleles which optimally code for the future existence of their phenotypes. Combative birds are more likely to survive in an environment of predation than pacifist birds. Let’s hope the balmy, isolated island did not fully select for pacifism - in the absence of threats (pacifism being usually biologically-cheaper). Or at least let’s hope that if you’re a bird; the rats take a different view.

The story is not much different in hunter-gatherer societies. For Malthusian reasons kin-groups invariably come into conflict with adjacent kin-groups who are less related. Genes which select for in-group solidarity and out-group xenophobia propagate more successfully: they breed winners.

Things changed with the advent of complex societies (initially agrarian). Cooperation groups became larger (such societies can support military/governance elites which work better at scale); social-solidarity then has to extend beyond the kin-group, beyond even the ‘family and friends’ of interpersonal reciprocal-altruism.

The in-group of coherent society is now more strongly defined by ideologies - religious/secular loyalties - underpinned by elite-obligations to all in-group members, cf. noblesse oblige. A society held together by cultural traditions which encode social protocols of mutual obligation and support is nevertheless more fragile than one based on blood and personal ties, more prone to break down into civil war and anarchy.

Capitalism is the current end-point of scaled-up societies. The transactional nature of large-scale production and exchange requires routine interaction between people who have little to no history of prior personal involvement. The operation of society - which rewards individuals in wages, dividends, rents, interest and above all commodities - is structured by processes of immense spatial and temporal extent.

This feeds back into the psychology of those who animate these processes in their own behaviours, elevating the propensity to prosociality. Often spoken of as an absolute good, a prosocial psychological makeup would have served most people poorly in prior societies, most specifically honour cultures. The prosocial default of generalised friendliness, tolerance of bad behaviour in others and a strong disinclination to violence would have marked the procial individual as a dupe, a coward and loser, who could be disregarded and perhaps dispatched with impunity.

But in recent centuries, the optimal psychology for making trade, and later capitalism work at the largest scales has indeed been prosociality. This combination of psychological traits has been strongly selected for in elites for hundreds of years now. The political corollary to prosociality is liberalism: it is no surprise that Western elites are by default liberal in outlook.

It has been noted that artificial people are, by design, highly optimised for the operations of global capitalism. The typical individual is prosocial in character, of high ability and meritocratic in outlook. Those are, of course, just the characteristics we find in high-functioning elites today. Models also present publicly as compassionate and charitable towards the masses of baseline people. This is not surprising as social cohesion is essential to their own continuing position.

Social dynamics: options

The biology of artificial people is currently shrouded in mystery. No public studies are available, genomic information has not been disclosed except to state that codons are non-standard as in synthetic biology. The reasons for secrecy are stated to be commercial confidentiality. If we assume, however, that artificial people will mate with each other to engender elite-level descendants then they will have psychological drives which encourage them to do so. In what follows we assume this to be the case.

How will baseline-elite individuals react to the novelty of a new and significant elite-fraction composed of a self-perpetuating community of artificial people? Baseline elites equally expect to have descendants and wish them well. We therefore expect baseline elites to wish to ensure they and their offspring are not discriminated against vis-a-vis artificial people, to therefore have ‘sharp elbows’ when it comes to scarce, contestable resources such as good schools and universities... but beyond that not to care too much.The size of the global elite which underpins the modern economy is far larger than any one elite member’s family and friends... and the global economy’s ongoing replication benefits every single one of the elite.

So in summary, provided the artificial people aren’t too ‘grabby’ - displacing baseline-elite members - the normal advantages of intra-elite prosociality will hold and the situation will remain stable within the elite.

What of non-elite baselines, the vast majority of humanity (the 99% as some have said)? 

They are used to seeing elites accumulating much of the material and cultural advantages of life - but providing they also see opportunities for themselves, their families and their friends, history has taught us that they are mostly reconciled to their lot. The ideologists of the media talk up meritocracy, persuading the masses that with fortune, talent and diligence success is always a possibility (you can become a pop star or a star sportsperson!) - and there’s always the lottery. Elite-sourced ideology in periods of stability can be encompassed by this simple slogan: ‘We are all in this together’.

It is not wealth inequalities per se which lead to social instabilities: all complex societies hitherto have been marked by such inequalities, often of grotesque proportions. Social instabilities occur when the mass of people actively suffer acute privation: new diseases, sudden job-loss, economic collapse, homelessness and/or elite predation.

Such instabilities spiral out of control when the elites themselves are riven by conflict. Factions strive to enlist and direct mass discontent in the services of their own ambitions. If their conditions of life turn bad, the masses will look in the first instance for someone to blame.

If the elite is structured by visible differences - race, colour, language and religion are all markers - then scapegoating is easier. Ugandan Asians, Malaysian Chinese and the Rohingya people in Myanmar are all examples. Bearing this in mind, and while they subsist as any kind of vulnerable minority, artificial people would be wise to display no markers of their artificiality, keeping the fact of such a closely guarded secret even in normal times - as periods of trouble can seldom be reliably predicted. And indeed we observe that people do not generally advertise the fact that they are artificial: a strong social taboo has arisen against even mentioning this possibility - or prying.

Baseline-elites will be strongly motivated to support artificial people if they come under hostile scrutiny. Class interests and solidarity will dominate. Elites are fully aware that attacks on one section of the elite - artificial people - could easily spiral out of control and generalise to attacks on all. Expect the ideologists of the media, elite-sponsored pressure groups and their supporters on the streets to be vociferous in their defence of artificial people against non-elite, ‘populist’ harassment.

But suppose the number of artificial people increases markedly, so that they inhabit more and more of the economy. Think of it as a wave of automation where the automation is by androids (the artificial people here are being considered as playing the same role as, say, humanoid robots). Perhaps the artificial people can be designed in variants: stronger; or more radiation-disease-pollution resistant; smarter; braver and so on. If it could be done no doubt it would be done.

The mass of baseline people would then perceive artificial people to be coming to displace them on merit: they are simply better. It doesn’t take a genius to predict the kind of protests and violence we generally term Luddism. Luddism to date has been manageable; displaced workers found better jobs. But in the scenario we are considering, infinitely-flexible artificial people are simply better than baselines at any job. (Consider artificial drudges for the dirty jobs, epsilons ridiculously happy to serve [3]).

To return to the question at hand: we can safely predict there would be resistance in this scenario, but would such resistance be right?

And again we answer: the question is wrongly posed in the abstract: right for whom?

If the non-elite baselines - the human masses - could be offered iron-clad guarantees that their survival/well-being and that of their families would be confirmed and that they would find (be provided with) fulfilling activities - then they would plausibly have an interest in being wholly displaced from the productive economy. In some sense all baseline humans would then have an opportunity to live like the unproductive elites of past societies such as those of antiquity. 

But without the power to enforce this outcome (necessarily by force in the last analysis) could this really be guaranteed? If artificial people had an interest in their own biological survival and that of their kind, then as an effective species they would have no vital interest in the ongoing existence of any baselines who were not necessary for the replication of their global economy.

To restate the argument. Imagine a global economy run by artificial people and serving in the first instance to replicate those artificial people. Why would this society have a vital interest supporting a parasitic caste of non-economically-productive baseline humans? Any such arrangements would have to be highly unstable. If you were such a baseline human, wouldn’t it be right to reject such a future, to prefer a suboptimal globalised economy rather than risk personal abandonment/extinction by an artificial-person-run superior version?

It might be objected that the mass incursion of artificial person variants into all the interstices of economic life is extremely unlikely precisely because it would be so incendiary. But such a dynamic is very attractive to capitalist economic and political elites just because - by hypothesis - it markedly raises the quality of ‘human capital’ while lowering costs. With such competitive advantages it would be hard to prevent such a trend from working itself through: after all, on any particular day it just looks like more automation improving productivity.

If the bulk of baseline humans really do become superfluous then perhaps they should just… vanish. Mass killing and expulsions would not be the way to go - way too risky and it offends against prosocial values. Much better if the baselines just stopped having children. The problem resolves itself then in just a few generations.

Designing artificial people who are rather better romantic partners than the average ‘baseline next door’ does not seem an impossible task. There is an extensive literature, after all, on what constitutes the ideal mate. Anyone who called out this tactic would hardly attract widespread support given the obvious upside to every single baseline if these romantic partners became widely available.

Another approach, perhaps technologically more speculative, would be to offer baselines the possibility to transform themselves into artificial people with all the consequent benefits of genomic improvement. The prospects of renewed vitality, better health and a longer life (potentially immortality) might well make this an attractive proposition for many. The issue of transfer of brain states encoding memories and identities would have to be resolved in this scenario: a challenging, but not insurmountable problem of neural nanotechnology.

One could also envisage compulsory, penal or covert transformations to neutralise key individuals or opponents of artificial people. Presumably once having been transformed, their loyalties might change: a shift which could in any case be tweaked during the transformation process.

In summary, from a baseline point of view, elite and non-elite alike, the key political question would be whether artificial people proposed to work with baseline humans in a cooperative way through shared institutions respecting democratic norms - or whether they considered themselves a countervailing agency seeking to supplant humans altogether and therefore with no commitment to existing institutions, laws or conventional morality. (This would in practice be a matter of dominant factions rather than the entire community of artificial people, who are unlikely to be a monolithic bloc). 

Again there are precedents for these distinctions. The revolutionary communist parties used to treat democracy (bourgeois democracy) as a train: once you have arrived at your destination (socialism) you get off; all’s fair in love and war. Theocratic Islamists have expressed similar sentiments.

Conclusions

If artificial people were simply non-reproducing ‘tools’ (instrumentum vocale [2]) without innate drives to reproduce themselves then they would not fundamentally threaten human existence.

If, however, they were designed as (effectively) a variant and more socially-competent species operating in accordance with their own species-interests, there could be a significant risk of baseline human extinction.

In the latter case ought we to care? The universe certainly doesn’t - species go extinct all the time, a consequence of the implacable workings of the laws of physics with complex boundary conditions [4]. 

For an individual whose subjective allegiance is to life expanding throughout the universe (a belief that more self-aware existence is always better) then artificial people as an optimized form of humanity for interstellar expansion is a compelling narrative. Unlike your direct kin, such people won’t be fabricated using parts of your own genome but so what? The caring emotions you feel for your kin (deeply of course) only exist courtesy of naturalistic evolution (brutally, any less-compelling feelings would have simply eliminated such defective genes from the gene pool). 

Artificial people make such feelings obsolete; design replaces natural selection. The truly far-sighted strategy is to rise above innate feelings about your own family, your own in-group and support those forms of life best adapted to a future of even greater social cooperation and unbounded scale. It’s the only way life is going to get off this planet and diversify throughout the galaxy. 

Individuals who take this view are more likely to have quasi-descendants in the remote future than those who don’t  - those traditionalists who will most likely have become extinct. And artificial people are likely to have a great deal genomically in common with human beings today, as baseline humans will constitute their basic template, their point of departure. One is reminded of the Neanderthals.

Is this ever going to be a popular, or even legitimate point of view amongst baselines? Probably not. There will be resistance to artificial people driven by the deepest emotions of self-preservation: people and their families. But in the longest time-frames the perpetuation of life like ours will be dependent upon artificial people winning out. So enlightened people should be on the side of artificial people even though it may well feel, and be perceived to be, disloyal.

A final point. Run time forward a thousand years. Earth’s ecosystem is dominated by artificial organisms, consciously designed using synthetic, tailored genomes, Robin Hanson [5] has argued explicitly for such a scenario. This post-natural civilisation has expanded to the stars. 

In a new solar system, it encounters a naturalistic baseline civilisation much like Earth’s back in the early twenty-first century. What should our post-human civilisation of artificial people plus designed organisms do?

Why not give our new contacts a makeover? We’d be doing them a favour - as well as gathering them harmoniously into our nascent interstellar civilisation...

References

[1] "Traditional and Critical Theory", Max Horkheimer, 1937.

[2] “De Re Rustica”, Marcus Terentius Varro, 37 BCE.

[3] “Brave New World”, Aldous Huxley, 1931.

[4] “La philosophie dans le boudoir, (Dialogue the Fifth)”, Marquis de Sade, 1795.

[5]. “Theories Of Unnatural Selection”, Robin Hanson, 2021.


Submitted: May 07, 2021

© Copyright 2021 AdamCarlton. All rights reserved.

  • Facebook
  • Twitter
  • Reddit
  • Pinterest
  • Invite

Add Your Comments:

Comments

hullabaloo22

A lot of food for thought in this, Adam.

Fri, May 7th, 2021 6:30pm

Author
Reply

It won't make a lot of sense unless you hauled yourself through 'La Double Inconstance' with its thirty chapters :(. You may notice a certain satirical edge under its dry, academic prose...

Fri, May 7th, 2021 12:07pm

88 fingers

When I saw the title and read the prologue, no offense, but I couldn't handle the technical aspect of your story.
I enjoyed your story, and after covid and if my girlfriend and I split, I might look into an AI for a companion.
Maybe get one that looks like Beyonce or a 70's version of Barbara Bach who was in the tv show, "The Dukes of Hazzard".

Fri, May 7th, 2021 10:49pm

Author
Reply

Ah, Daisy Duke...

Sat, May 8th, 2021 1:14am

Stellanotte

Chomsky's dismissal of the French post-structuralists aside, one might ask oneself, 'What would Chomsky say?' https://www.theatlantic.com/technology/archive/2012/11/noam-chomsky-on-where-artificial-intelligence-went-wrong/261637/

Tue, May 11th, 2021 5:45pm

Author
Reply

AI is my main area of expertise and I've always been a huge fan of David Marr and his multi-levels of analysis. People correctly observe that 'deep learning' - artificial neural nets - is today engineering, not science. However GOFAI is an equally-inadequate paradigm in confronting the phenomena. Feynman used to say that until you can build it, you don't understand it, and things like cognition, emotion and even basic, brain-stem-type systems control will continue to be mysteries until we're building systems whose design specs require such mental phenomena. With Google, DeepMind and all the other guys at present we're in the merest foothills of the journey, clutching at the low-hanging fruit of massive Internet datasets, looking under the lamp post. It will be better in the decades to come, particularly with non-trivial situated automata out there in the world, with social skills. It will be hard -remember the cyberbaby!

For the record, I think that Chomsky is pretty much right in every answer he gives in this interview you cite.

Tue, May 11th, 2021 11:50am

Stellanotte

Similar to Feynman, Seymour Papert proposes Constructionism, learning through simultaneously building material objects as well as knowledge structures. It's fun to review the good old days of my Linguistic and pedagogical studies, and I wish I were clever enough to use all of that to write an equally satirical rebuttal. Maybe when I retire one day ;-)

Thu, May 13th, 2021 5:27pm

Author
Reply

Writing critiques is, unfortunately, the easy part. The scientific study of mental phenomena (emotions, intelligence, social deftness, etc, etc) is the hardest research programme humanity has ever attempted. Hence our lamentable efforts in building chatbots, emulating call-centre staff and automating (for example) care home staff work.

But it's really interesting, isn't it. You write your stuff and I won't be the only one reading it!

Thu, May 13th, 2021 12:53pm

Facebook Comments

Other Content by AdamCarlton

Short Story / Thrillers

Short Story / Science Fiction

Book / Science Fiction