AI labs' job-loss plans won't say what the public gets
7 hours ago
- Frontier AI labs like Anthropic are publishing economic policy papers on labor displacement, proposing tiered responses to rising unemployment from 5% to catastrophic levels.
- Anthropic's framework includes wage insurance and job matching at 5% unemployment, extended benefits at 10%, and basic income or equity sharing at the highest tier, but avoids specifying the public's share of capital.
- The author criticizes this as 'preparedness theater,' where institutions build elaborate plans without addressing the simple distribution question: how much capital income should belong to the public.
- Matt Bruenig argues AI displacement is not new; the solution remains universal public ownership of capital, like a social wealth fund with a universal dividend.
- DeepMind's paper uses unorthodox methods like simulated economists to score policies but also omits the size of the public stake, recommending waiting until data shows labor income collapse.
- The literature contrasts speculative ideas like 'Whuffie' (reputation currency) with reluctance to endorse strong democratic claims on capital, highlighting a political bottleneck rather than a technical one.