Summary
Bloomberg Opinion's Parmy Olson discusses the recent calls by AI leaders to slow down development due to safety and existential risks. She argues that despite these warnings, a true slowdown is unlikely as US companies are locked in a competitive race with China to achieve superintelligence first.
- AI leaders like Dario Amodei have expressed genuine concerns about the existential risks of superintelligent AI.
- Silicon Valley's ideological belief that superintelligence is inevitable drives a race to build it first.
- US AI companies are highly concerned about China copying their models through a technique called distillation.
- Olson predicts there will be no actual slowdown in frontier AI research, though independent auditors may be introduced.
- Unintended consequences of AI could mirror the unforeseen societal impacts previously seen with social media platforms.