Being Distilled by AI Is Our Fate! And Then What?
This year, with the craze around Crayfish and skills, everyone is getting more anxious: write your work experience up as a skill, and then the boss uses AI to replace the working drudge. This is sitting at a critical point, and it’s no wonder people are anxious.
In fact, AI “distilling” humans is already very mature in some professions. Translation, for example: the bilingual material human translators contributed has all become AI’s feed. Painting, too — artists’ pictures have been eaten and digested by AI. Programmers even more so: it is the profession most fond of sharing its own output, and all that open-source code has ended up as the programmers’ own gravedigger.
What has people more panicked right now is simply that AI has got hold of more and more tools, its capability expanding rapidly from conversation to operating the digital world and even the physical one. When you watch a sewing worker wearing a camera to record their own movements, feeding a robot model that will replace them, do you find that terrifying?
The fear of being distilled by AI comes from two places.
The lag in wealth redistribution
The fear that AI replacement brings comes first from distribution. Overall, AI raising productivity must be a good thing. Thought experiment: suppose your monthly salary is 10,000, and you quietly use an AI stand-in to do your job at a cost of 1,000. Then you can keep drawing 9,000 a month and never go to work at all. Would you take it? I think almost everyone says yes — there really aren’t that many people who love going to work. The problem is your boss knows this too, and he doesn’t want a middleman pocketing the spread, so you lose your job.
A sharp drop in labor cost, through market competition, drives a sharp drop in product prices. But with everyone replaced by AI, no jobs, no wages, no money to buy products, supply ends up massively overproducing.
Under the traditional distribution model, allocation basically follows scarcity: capital, technology, labor all take part. So who should AI’s output belong to?
The people who invent, build, and maintain AI must get a share. But AI has a feature no previous technology ever had: it was produced by extracting all of humanity’s knowledge, so it ought to give back to all of humanity, and must never become the private instrument of some person or some group. The scenario in Liu Cixin’s novella Supporting Humanity, where the “Final Owner” alone holds all of humanity’s property, is something no person with a conscience wants to see.
So the redistribution of AI wealth is something every human being should fight for.
Rethinking the value of labor
Suppose the redistribution mechanism is solved, and AI supports humanity well. We still face an inescapable question: what is a person’s value, as such? When we were still working for money, this question was hidden by the day-after-day cycle of earning and spending. At work, do we feel our work has value? Do we actually like our work? Would we still want to do it if it paid us nothing?
The marathon began as genuinely valuable labor in real military struggle, and that value was long ago replaced by modern transport and communications — yet many people still love distance running. In Go, humans have completely lost any hope of beating AI, yet many people still love playing Go. Marathon runners and professional Go players point to something: perhaps what everyone ultimately pursues is self-improvement, while being a useful person in the utilitarian sense becomes less and less realistic.
So-called financial freedom has two directions, earning and spending. By lowering your desires, maybe we don’t have to wait for AI to support humanity — a lot of people can in fact already start thinking: when earning money is no longer necessary, apart from consuming, what else can we do?