With $7 trillion on the line, AI safety may always take a backseat to market domination
Why slowing down AI scaling may be impossible.
TL;DR
- The race for AI supremacy may be impossible to reconcile with AI safety efforts due to financial incentives.
- Top AI companies propose independent oversight while seeking an AI slowdown, but critics argue trillions of dollars in incentives will prioritize progress.
- The complexity of AI systems and the deployment of autonomous agents make real-time auditing extremely difficult.
- A trust gap exists for independent safety groups due to industry ties and shared investors.
- OpenAI and Anthropic face pressure to deliver better models to justify multi-trillion-dollar valuations.
- Over $7 trillion is expected to be spent on scaling AI in the next five years.
- Daniel Kokotajlo, a former OpenAI researcher, believes AI cannot be safely scaled and suggests companies should avoid model improvement to force a slowdown.
- Some, like President Trump, see a safe future for AI and do not want to stifle its growth.
- Akshay Krishnaswamy suggests open models as a way to achieve safe scaling.