Post-instrumentality

I was reading Deep Utopia by Nick Bostrom a while back, I found him circling around the subject of "post-instrumentality". It means a time when all tasks we set out to do can be done better by something else, whether it is AI or other superhuman beings. Not just some tasks, but all tasks, he gives a lot of examples from work to parenting. In this kind of world one can lose the purpose of doing things.

We can call this hypothetical condition, in which we have no instrumental reasons for doing anything, the age of post-instrumentality. As we move toward this weightless condition, blasting away from the gravitational pull of the ground and its tough “sweat of the brow” imperatives on our days and our strength, we may begin to feel an alienating sense of purposelessness, an unanchored “lightness of being”. We are left to deal with the discovery that the place of maximal freedom is actually a void. [Deep Utopia, Nick Bostrom]

This kind of world doesn't come instantly, and we are starting to see hints of it already. Arguably parts of this have been true for many people even without AI. If you set out to do some task, there is usually someone who does it better, they are just not available to do the task. But is there something unique when your job is replaced by machines you can't compete with?

Lately mathematicians have been debating what their purpose is, if machines do the proofs. Architects and 3D modelers are now questioning their place after the release of GPT-6 Astra. Of course before any of these, we had software engineers grappling with agents writing code in superhuman speeds, if not with the quality.

Many solutions to purpose of oneself appear to coalesce to a spectrum. The anthropocentric old worldview, where we keep trying to cling to old ways of doing things, even by banning AI altogether. The other end of the spectrum is the transhumanist worldview, where we seek to use the tools to advance human condition, and eventually reach a post-humanity by merging with the machines. Even if I were to adhere to more transhumanist viewpoint, I don't think it is something that can be forced upon everybody.

Bill Gates recently presented his solution of "human reserved" jobs in his post the turbulent AI era is here. The choices we make now are critical. He gave an example of a care giver, and went as far as suggesting some menial jobs being part of it:

We might set something aside as Human Reserved for economic reasons. For example, we may do it because allowing machines to take over a certain role will displace a large number of people who can’t easily change jobs. You can’t tell a 55-year-old who has worked in construction their whole career that they need to go work at an elder care facility and expect them to find it fulfilling.

Not only is this unworkable, but it is likely unwanted in the big picture. Most people would not choose to go stock shelves at 4am, or climb to a construction crane if given a choice. Yes there are a few people who would keep doing it for recreational reasons, but most would not. Additionally, if autonomous robots can do these tasks better, incorporating humans into a fully automated construction process would be a hindrance.

There is a deeper question here, why does he worry about retaining construction jobs? If you read between the lines, he sees the act of working as a meaningful thing for the worker, in other words, he is unable to imagine communities of other kinds replacing it. He doesn't seem to be proposing anything replacing jobs rather than jobs.

Similarly Mark Zuckerberg in his post The Future is for Everyone paints a picture where superintelligence is a mere tool, to improve ourselves and the work we do. I must admit I like his enthusiasm for helping people build great businesses and fulfill themselves, even though it is misplaced. He also seems to misunderstand how capitalism works:

I do not understand why anyone who believes that AI will eliminate most jobs and much of humanity’s relevance would rush to build that future. [Mark Zuckerberg]

In capitalism, efficiency is what businesses are after. If a company can replace a worker with a more efficient robot, they will do so. Maybe he doesn't see AI replacing all the jobs, but he keeps mentioning superintelligence 61 times. He sees it as some great equalizer, as he continues like this:

We believe that delivering superintelligence to everyone is the way to answer this question. This follows the tradition of putting the power of supercomputers and the internet in everyone’s pockets and on our desks. Rather than centralizing superintelligence, we should distribute it widely and give every person the ability to direct it. [Mark Zuckerberg]

That is like dogs conspiring to have good human hosts. Superintelligence by definition is more capable than us, it will be able to answer all questions we have, better than us. It won't have answers to all questions, but it will certainly run business better than any human being. What I think happened is that Zuckerberg has mixed terminology, and is not in fact speaking about superintelligence but some dumb agents.

I've also seen other proposals for purpose in the post-instrumental world. One view I saw is that we will develop a "leisure economy" where people will be left to make fantastic new worlds for other people, and somehow it would still retain the capitalistic concept of money for work. I don't see how this would happen if AI is more capable than us, because it can also create these worlds better than us.

Money would exist for a long time in human cultures, even though its meaning would change. This is because a human born into the post-instrumental world has no skills or abilities, other than their relation to other people, to attain money. In this kind of world money as a concept would end up representing 'appreciation' among people. After all somehow we must choose who can take part in a theater premiere.

All of this depends on how deep in the future you go, Nick Bostrom covers some very utopian views, which I don't find useful to think here. For instance, one could imagine atomically precise reconstructions, which would of course mean that everybody can take any seat in the premiere, because there could be infinitely many theaters and seating arrangements.

For those in the anthropocentric group, they have a clear solution, ban the AI and there will be no post-instrumental world. For those of us who don't adhere to that, there simply is no solution in human form, purpose is bound to lose its meaning. It doesn't mean the time we spend as humans will be uninteresting, after all most of us aren't working in jobs where the primary motivation is the prestige of accomplishment. It might very well be that people who seek prestige are most hit by these questions, as they are less likely to achieve it.

The upshot is that the signaling hypothesis paints a relatively rosy picture of our prospects for being able to sustain high levels of interestingness even if we live for a very long time. One could in principle keep yachting, or listening to chamber music performances, or taking gallops around the grounds, or attending galas with finely dressed people—in perpetuity, if such were one’s fancy. [Nick Bostrom, Deep Utopia]

Nick Bostrom's book Deep Utopia is worth a read, even if it is sometimes a bit longwinded, if you are already accustomed to the topic.