Anthropic says it blocks users seeking biological-weapons research help
Summary
Anthropic says it has blocked several attempts to use its AI for biological-weapons research, while former safety leaders Evan Hubinger and Paul Christiano warn that the industry still lacks an adequate plan to control potentially catastrophic superintelligent systems.
Key Points
- Evan Hubinger estimates AI has more than a 10% chance of killing all humans within the next decade and says Anthropic lacks a plan to solve superintelligence alignment.
- Paul Christiano says the AI industry, including OpenAI, is not on track to reduce catastrophic, irreversible loss-of-control risk to an acceptable level.
- Anthropic says it thwarts several users' attempts to conduct research that could help develop biological weapons.