Anthropic co-founder warns: artificial intelligence needs a 'brake pedal'

Anthropic co-founder Jack Clark warned on BBC's Newsnight that artificial intelligence could reach a point where it develops without human intervention. Clark stressed the need to create mechanisms to keep AI development under control.

Jack Clark, co-founder of the artificial intelligence company Anthropic, made a concerning warning on BBC's Newsnight: artificial intelligence needs a so-called brake pedal that would allow humans to halt its development if necessary.

Clark explained that artificial intelligence could reach a stage of development where it improves itself without direct human input. This would mean that humanity loses control over one of the most powerful technologies ever created.

Anthropic is one of the world's leading artificial intelligence research companies, and has itself emphasised the importance of safety in AI development. The company is known for its AI assistant Claude, which is considered one of ChatGPT's main competitors.

Clark's remarks reflect a broader growing concern among artificial intelligence experts that the field is developing faster than legislators and society can adapt to it. Several prominent technologists have called for international cooperation to place constraints on AI development.

The topic of artificial intelligence regulation has come to the forefront in recent years both in the EU, where the world's first AI regulatory law was adopted, and in the United States, where the discussion is still ongoing.