A new P-Zero Research benchmark measures how well AI models design and assess experiments on their own. Scores are rising so quickly that Jack Clark now believes self-improving AI could arrive by 2027.
In Foreign Affairs, Niall Ferguson, Lisa Klaassen and Eric Schmidt describe the US and China as AI empires and other countries as dependencies, arguing the race will be won by whoever can pay for it. US companies alone plan to spend over $600 billion in 2026, though debt, data center protests and a possible Taiwan chip crisis could slow both powers.
Some researchers believe recursive self-improvement could trigger an explosion in AI capabilities and in the associated risks. Others dispute that outlook and consider the automation of research manageable.
Sam Altman intends for OpenAI to develop its own humanoid robot, and a demonstration could come next year. He stops short of promising that a commercial product will follow soon after.
Sam Altman outlined to OpenAI staff three possible endings to the global AI race and called the third one terrible. 01net notes that this scenario looks a lot like his own company's path.
Mistral chief executive Arthur Mensch warned on October 20 that Europe would be left with 'leftovers' if it stopped building frontier AI models. Relying on apps built on American or Chinese models would also create a big trade balance problem, he said.