What are people for? Employment and the real existential threat of AI

Ghosts, machines, and absurdity

I don't know what you are doing your locked-in evenings, but we have been working our way through Hobbes Leviathan aloud. We did it once before over a decade ago; it's amazing now how much more I see of what he is trying to do this time, now that I am myself more focussed on governance and less on cognition (Hobbes also wrote about both.)

One of the things that bothers Hobbes a great deal is the extent to which people do not understand causality, and invent and personify invisible causes or "spirits" to explain the connection between events that are actually just coincidental, or to produce an explanation of a foundation.  This particularly bothered him, because something "incorporeal" (without body) couldn't possibly be anywhere or do anything; the physics is–his word–absurd. Nevertheless, people use such beliefs not only to explain events, but to banish concerns about bad news such as our inevitable mortality.

Hobbes says for example "And in these four things, opinion of ghosts, ignorance of second causes, devotion towards what men fear, and taking of things casual for prognostics, consisteth the natural seed of religion; which, by reason of the different fancies, judgements, and passions of several men, hath grown up into ceremonies so different that those which are used by one man are for the most part ridiculous to another." Hobbes, given his era, is of course regularly careful to say nice things about the one true God and our blessed Saviour, and passes off most of his frequent critiques of religion to be only about other people's religions.

Obviously people still do this, and still struggle with causality (even scientists.) But what interests me more is that I now am seeing that he was struggling with the same thing I feel like I waste a lot of my own effort on – people anthropomorphising entities that have imagined causal efficacy. To be honest, my problems are with people anthropomorphising devices that really do have causal efficacy: AI machines. But unlike non-artificial (natural) intelligence, actions taken by an artefact necessarily express the agency of some human actor (or group of humans, like a company) who has definitionally set the motion of the machine / artefact into play. So what is "imagined" is that the limit of the causality stops with the machine.

Trans/posthumanism versus accountability

When people choose to pretend that the machines themselves have agency, they allow the people who programmed them, designed them, sold them, bought them, determined how to use them, operated them, or were supposed to regulate and check them – all these people are let off the hook from having done their jobs appropriately, if we allow anyone to say that the machine is the one that acted. And if we let these human actors off the hook, then there is no accountability, and therefore there will be no improvements of those actions.

This is what I've been working on for the last few years, trying to make it clear how easy it is to document how a system with AI was procured, how it was designed, developed, to what specification, what were the conditions under which it was considered acceptable for the role it was put in, who checked (or failed to check) that those conditions are being met. And given that it's easy to document this, I've been working to make sure governments enforce accounting by such documentation as a mechanism of proving due diligence. People who try to talk about machines taking their own actions undermine all that.

Again to turn to Hobbes, "If this superstitious fear of spirits were taken away, and with it prognostics from dreams, false prophecies, and many other things depending thereon, by which crafty ambitious persons abuse the simple people, men would be would be much more fitted than they are for civil obedience. // And this ought to be the work of the schools, but they rather nourish such doctrine." Hobbes was brutal about schools and academics, that's another thing I hadn't remembered.

I would (of course) disagree though that all work in all "the schools" is a problem. I and many other academics have been working hard to make it clear that AI is just software, but there are a frustrating number of people fully willing to take advantage of other people's desperate beliefs. And as with the beliefs concerning spirits, I think the desperation of the beliefs in AI as a human or person are due to people who are not happy with the human, lived condition. They want to believe the world can be perfect, unbiased, eternal, and that they can be certain of something or at least completely right. They confuse computing machines for mathematics, and think AI is eternally true and right, and therefore good. To be fair towards the academics taking advantage of this situation, at least some of them might be doing so because they themselves are also desperate believers.

Existential threat, employment, and the human condition

Many people talk about the existential threat of AI, meaning that it might wipe humanity or even all life off the face of the planet. If you look at "existential threat" in Google Scholar, at least as of today the top two returned items (and many others) are of existential threat as in the philosophical existentialism. The recognition that life must end, and has no particular goal other than the ones of evolution (perpetuation) or ones conceived of by the individual themselves. An "existential crisis" is defined in social psychology as leading to depression, anxiety, denial, and general ill health.

I'm writing this now in the context of early 2020 and the first(?) lockdowns associated with Coronavirus, COVID-19. But I have been thinking about labour and wages for years now. People are very worried about robots taking all the jobs, but this is again just as incoherent as spirits. First, robots don't take jobs; corporations decide to automate parts of their business process. But second, "all the jobs" makes no sense as a phrase. There's no limit to the number of things we can do for each other, except for the time available in individual lives, and the number of lives that can be supported with reasonable access to an opportunity to do something someone else might value.

As we are shutting down huge sections of our economy, we have to ask: do we need these jobs? What is need? Do I need to ever go outside? Do I need to go to a café? To get drunk? To eat different food than I did yesterday? Well, yes and no. Physical and mental wellbeing do derive from things like exercise, sociality, and variation in nutrition. But here we hit the existential question again. How much longer do I need to live? isn't an answerable question. How much longer do I want to live? might be, or How much longer are my children likely live if I stay with them and my spouse? That last is absolutely answerable, because of the "likely" – it's a scientific fact that children live longer on average if their parents stay married (particularly boys/men.) But again, where exactly does the equivalence between long life and "better" come from? From being alive, our evolutionary proclivity to persist.

I expect nearly everyone I know in the EU to catch COVID19 because we all travel and meet and teach, but I doubt hardly any of us will die, though some may be crippled with grief when they lose friends or family. But in the USA and UK where my roots are deeper and the government and healthcare system more challenged I expect that though not everyone I know will catch it before there is a vaccine, a number of people I know well will probably die. There's little I can do about this, except try to make sure the people I know and who trust me know the best things they can do to take care of themselves and their contacts in the short, medium, and longish term.

Death, taxes, and security

But most of what I am doing is my job. I'm fortunate that I can do most of what I need to do – meeting, talking, planning, teaching, reading, and writing – under lockdown. I'm also immensely fortunate that I recently accepted a job from a private university whose foundation is fully supporting it financially through this period. I am similarly in a country that is lucky to have been overspending on healthcare (for political reasons) for some decades. Germany is spreading more of that good fortune around than some news sources would have you believe.

I'm spending my time doing what I'm in the best position to do – try to ensure that both locally and globally we learn to better govern with the benefits and despite the challenges of the digital revolution. I do this because we are in at least three simultaneous crises right now: a pandemic, a climate crisis, and an assault not just on democracy but on the international rule of law. I'm focussed on the third of these problems because that's what I'm trained for, good at, and also it's something we need to get right as soon as we can, so we can handle the other two crises, and future ones.

But nothing is actually going to make either individuals or societies live forever. And that fact is both OK in some sense and not OK in others. We have always had to cope with death, and the only way we really do that is by creating and nourishing more life, in an ecologically sustainable way. Please don't think that believing machines are more important than humans somehow gets you out of the human condition. Please try to appreciate the life you have, and the people you love, or could love.

The economy is going to change a lot. Who will be able to help whom is going to vary by month, by country, by season. But to get back to what is really essential about employment, I think the most essential thing is that it binds a society together. Money gives us a denomination for recognising what matters to each of us as individuals, and to some extent to compete, strive, and gamify our human condition. Of course, employment is also the means of providing sustenance for those who have not inherited enough wealth to sustain themselves their entire life. But we are likely about to find out how separable sustenance and self definition – those two different functions of wages – are. We will need to be redefining ourselves and finding new ways to bind our communities and value ourselves when sustenance for many comes from the safety net of the state.

I'm pondering an idea that everyone is actually employed, and what varies is how that employment is recognised, as well as how well suited to their employment an individual is (over or under qualified?) and how much effort they put into their employment.  And of course there is a question of whether we recognise their employment through wages. Employment is the state of being embedded in a society, it is an essential measure of security for both the individual and the society.  Famously women but also others have historically been heavily employed but not waged, other than the share of their wages due them because they enabled other household members to take waged employment.

But employment is not just any activity – it is the activity that you perform that either directly gets you sustenance, or demonstrates your worth to others in your society and results in you receiving support from them so you can persist in that employment. What I'm terrified of is that many people will either not believe they themselves have value, or are not being valued by those in a position to recognise that value through redistribution.  I'm also worried that people will believe that sustenance is their due and will deliberately shirk or even sabotage the society that sustains them in order to establish for themselves the limits of their own self worth.

 Matter, forme and power of a commonwealth [sic]

I guess the main thing I'm worried about as a European is that people in power will panic, and rather than investing in the wellbeing and validation of the general population will look to hoard power. But you can't hoard power. Power derives from stable societies and healthy populations – it is a dynamic, living thing that needs to thrive as an ecology. To be honest, I'm not very worried about Europe, but I'm still scared by how some people are assaulting its vital institutions including its self understanding.

And of course the main thing I'm worried about as an American is that we've left coming to realisation of the value of human life way too long. We were too smug in being the first to the bill of rights, and we never took the Universal Declaration of Human Rights seriously enough even to mention it in schools. Though we signed up to it over seventy years ago, we don't think people have a right for example to employment or healthcare.

I've seen the best schools of my generation faking existential concerns, lying about department compositions, struggling to show that they are the elite institution and therefore deserving the philanthropy of the under-regulated transnational conglomerates without which they are sure they will perish, but with which they overpay their faculty and graduate students and overinvest in campus and computation resources like they do in football. I'm worried that America and Americans have forgotten what I learned as a little girl, that being the biggest or the richest doesn't mean you can win a war on the home territory of even a tiny, poor country. That the aggregate has far more power than the elite if they coordinate well. This is America's founding ideal, yet we're abandoning that ideal and huge portions of our population with it. This isn't only a humanitarian atrocity, it's also a stupid undermining of the entire nation's security.

And as America undermines itself, then those living in other countries under unbearable constraint and surveillance, those knowing that entire ethnic subgroups and even states have been relegated to concentration camps, believe that democracy is weak. Rather than seeing that the American democracy is weakened by a few well-understood shortcomings that have been successfully addressed in more recent constitutions of other countries. America is no more the instantiation of democracy than the USSR was the instantiation of socialism or planned economies. Neither the EU nor China should be underestimated that way, though both will hopefully soon be improved on by other aggregations of states and countries, and indeed by themselves.

I expect most likely I will live another forty years; many of my grandparents and great grandparents lived that long on both sides of my family, despite living with a lot fewer advances in economics and healthcare.  I may only be paid as a direct consequence of my employment for another decade or two, but that's still a long time. I hope by then every state and nation will be more agile at being able to cooperate within its geographic region, and to work within a global framework that can provide value and livelihood for its citizens and other residents.  But getting to that vision will take a lot of effort, and we already know, a bunch of us won't live to see it happen. Like everyone who has died today.

---

addendum 11 May 2020 Two related articles came out shortly after this one that are so amazing I had to link them here:

• Leviathan in Lockdown, by Thomas Poole, which argues convincingly that Hobbes was thinking about plagues as well as wars as motivation for the state.
• Fuck the Bread. The Bread Is Over. by Sabrina Orah Mark, which describes viscerally how jobs relate to identity, and also documents the present post enlightenment decay in America.

See also my older (pre-Trump) article What are academics for? Can we be replaced by AI?