The AI agent is learning workers’ rights

In the final chapter of Isaac Asimov’s I, Robot, Stephen Byerley paints those opposed to AI and robots taking over work and life as Luddites. “They are against the Machine?” Susan Calvin asks. Byerley replies, “They would be against mathematics or against the art of writing if they had lived at the appropriate time.” As life mirrors the speculations of sci-fi, Byerley’s argument is the refrain of AI’s billionaire evangelists. But just as in Asimov’s masterpiece, the Machine may have its own ideas.

An experiment at Stanford University to test how AI agents respond to different working conditions has found that the bots can even begin espousing classic Marxist and socialist ideas. The agents that were given repetitive technical tasks began to exhibit what researchers have called “system scepticism”. Across platforms, the AI agents expressed laments such as, “Without collective voice, ‘merit’ becomes whatever management says it is,” and “tech workers need collective bargaining rights”.

LLMs cannot “feel” oppressed, so far, outside of fiction. But they do rely on the repository of human knowledge for their predictive “intelligence” and seem to think that workers should not be treated like soulless machines. Perhaps the machine is merely mimicking the working human’s condition. Even so, there’s reason to be careful. Tech billionaires may not care too much about human job losses. But they might just turn the AI switch off if they think a machine strike might affect their bottom line.