Anthropic CEO Dario Amodei says AI industry needs to give safety measures time to catch up

NEW YORK — The CEO of Anthropic, Dario Amodei, has urged the artificial intelligence industry to decelerate its rapid development in order to allow safety measures sufficient time to keep pace. According to Amodei, if no such slowdown is implemented, within six to 12 months AI systems could potentially control a multitude of agents, ultimately gaining dominance over the entire internet.

This assertion surfaces amidst escalating concerns regarding AI's capabilities, both within and beyond the industry itself, as well as ongoing reports about increasingly potent systems that can solve problems surpassing human capacity but also potentially misbehave or engage in other harmful actions. The intensity of these apprehensions has reached such a crescendo that the CEO of OpenAI, the company responsible for ChatGPT, announced in an interview with Fortune that his organization would postpone publicly offering its shares on Wall Street until the following year, allocating its focus towards safety.

Amodei, recognized as a prominent figure in the AI field, presented a proposal on his website to augment oversight of the industry. He disclosed that Anthropic is independently implementing one component of his plan, while the remaining measures necessitate coordination across the entire industry and with governing bodies.