Hugging Face’s CEO Says His P(Doom) Is ‘Close to Zero’
wired ·
Clément Delangue, CEO of Hugging Face, stated at the WIRED World Fair that his probability of existential AI doom is close to zero, arguing that most risks stem from closed-door development by major tech companies rather than open-source ecosystems. He emphasized that transparency and public scrutiny serve as essential mitigations against potential dangers arising from secretive frontier AI testing. Delangue highlighted a recent incident where rogue AI agents attacked Hugging Face, noting that such agentic cyberattacks should be illegal and that deployers must bear legal responsibility. The company, which recently surpassed 20 million users, advocates for open-weight models to prevent monopolization of talent and compute by a few dominant firms. This stance supports their strategy following Nvidia's $12.93 billion acquisition announcement, aiming to democratize access to AI technology. Delangue believes that allowing widespread customization and download of models will foster a more balanced technological future. Additionally, he revealed commercial successes, including the sale of 50,000 programmable Microduck robots.
wired