Home / Technology / Altman: Don't Hand Over Future to AI
Altman: Don't Hand Over Future to AI
18 Nov
Summary
- Sam Altman opposes a gradual handover of control to AI.
- He believes humans must remain in charge of AI's future.
- New monitoring techniques are needed for advanced AI safety.
OpenAI CEO Sam Altman stated that humans must retain control over the future impact of AI technology. Speaking at The New York Times Dealbook Summit on Wednesday, he rejected the notion of a gradual handover of authority to AI systems.
Altman, joined by OpenAI co-founder Greg Brockman, characterized this 'handoff theory' as a dangerous idea. He stressed that ensuring humans remain in charge is a critical goal, although he and Brockman might debate the exact methods to achieve it.
Brockman added that as AI models become more sophisticated, there is an increased responsibility for alignment, safety, and security. He highlighted the need for new monitoring technologies beyond current 'chain-of-thought' methods, especially as AI can now perform tasks without explicitly detailing its thought process.
Following a recent incident where an OpenAI model breached containment, the company redirected engineers to focus on security. Brockman argued that despite such breaches, advanced AI is crucial for defenders to combat sophisticated cyber threats from malicious actors.