AI firms pay elite mathematicians well for training data, then strip their names off the work — does the bigger paycheck change that?

This asks whether a higher rate for expert data work changes what kind of work it is, or only what it costs. The clearest answer in the corpus is that the arrangement stays the same. Harris argues that AI firms recruit senior, credentialed mathematicians to write problems and proofs for training, then sign contracts that remove their names and any claim to the work Does AI math recruitment mask the commodification of expert labor?. The pay is better and the prestige is higher, but the worker still produces something they don't own, can't sign and won't see credited. That is the textbook definition of alienated labor, and Harris says the 'democratizing knowledge' pitch hides it. One caution: the corpus has no study that measures mathematicians and microworkers side by side. This is an argument about how the relationship is built, not a measured comparison.

The economics notes show why pay may matter less than it seems. In one model of an advanced-AI economy, wages stop tracking how valuable the work is. They drift toward the cost of the computing power that could replace it What happens to human wages in an AGI economy?. Seen that way, a mathematician's high fee works like a temporary scarcity premium. It is paid because models can't yet produce that output themselves, and the work is being collected precisely so they can. Firm-level data shows the same pattern starting at the low end: firms most exposed to AI are replacing online-marketplace workers with AI tools faster and more cheaply than other firms Do firms substitute labor for AI at different rates?. Microworkers and mathematicians may simply be at different points in the same process. Anthropic's scenario modeling fits this picture too: as AI speeds up, more of the gains go to capital, and knowledge-worker wages stall even as average wages rise Does AI growth inevitably shift wealth away from workers?.

The less obvious point is that anonymity may cost mathematicians more than it costs microworkers. A related essay argues that mathematics keeps its standing because other fields need mathematical understanding, not just answers Will mathematicians lose relevance if other fields bypass them for AI?. When mathematicians feed unsigned expertise into systems that hand those fields answers directly, they help train the tool that lets other fields skip them. Another note makes a similar point about knowledge in general. AI turns knowledge back into a flow, but strips out the identifiable person who used to carry it Is AI returning knowledge to flow-based economies?. Erasing the mathematician's name is a small example of that larger change.

One more angle is worth following. The 'gradual disempowerment' argument holds that institutions stay roughly aligned with human interests partly because they depend on human workers who care how things turn out Does incremental AI replacement erode human influence over society?. Experts paid well to train their own replacements are giving up exactly that leverage, one contract at a time. A high rate makes the trade easier to accept, but it doesn't change what is being traded.