Matthew Mandel

I’m very interested in what society looks like if we achieve AGI but don’t end up in a totally unrecognizable utopia/dystopia. One question is what work will be meaningful in such a world. A lot of what we do today will be rendered irrelevant, but I’ve developed some conviction that at least three professions will continue to be important.

No, the oldest profession doesn’t make the cut.

Politician: For a hundred years or so, political theorists have debated the role of expertise in politics. The modern state balances deference to technocrats and democratic accountability, but the question of how to balance them is still very much alive (see the fate of Humphrey’s Executor). When we have AI experts orders of magnitude smarter than the smartest humans, will the scales totally move towards AI-led technocracy? I don’t think so. Supergenius AIs might replace much of the bureaucracy responsible for figuring out how to achieve our ends. But we’ll still want human politicians to participate in a democratic process to determine what those ends are. Even if you’re a moral realist and think AI will be able to deduce the correct ends to pursue, morality leaves many questions indeterminate, requiring humans to figure out together what to prioritize and how to weigh considerations. The idiosyncrasy of our interests means we’ll continue to value humans who effectively and responsively aggregate and represent them.

Athlete: People like to point out that chess has gotten more popular post AlphaZero. That’s probably just a function of streaming personalities becoming a thing (I’d assume good chess players would find the scoresheet of an AI-vs-AI game more interesting than that of a human game, but maybe not?). Chess’s fate aside, human excellence in physical domains will definitely remain interesting to us. We’ve never cared about athletes because of their absolute ability but because of what they can do with the same machinery the rest of us also have. Sports are interesting because of human constraint, not despite it. If we’re not more interested in greyhound races where the dogs run more quickly than humans possibly could, we wouldn’t care more about robot athletes that also objectively outperform humans. Until biotech progresses to a point where we don’t have shared physical constraints, we’ll continue to (and potentially increasingly) prize athletes and physical competition.

Clergy: The future I’m considering is very weird. It would probably be the weirdest thing any human in history has experienced. And it will provoke existential questions about what it means to be a human, what is ultimately valuable, and so on. Religion becomes much more important in such a world. Religion is already making a comeback, throwing into question end-of-history-2000s-atheist-style arguments that religion is on its way out. You could imagine a Third Great Awakening leading to the creation of new religions that might explicitly situate AI in their theology. But I think incumbent religions will persist, and they have rituals that can only be done by humans (eg consecrating the Eucharist). The more things change, the tighter we’ll cling to old traditions, which often require the work of humans. Being ensouled, it turns out, is a competitive advantage.

Oh, and venture capitalist, of course.