Opinion: Are AI Labs Responsible Enough to Manage the Risks of Machine Intelligence?
joshua_saxe · x · 2026-09-02
Arthur Callis argues that AI labs are, in absolute terms, more constitutionally responsible than most industries like chemicals, energy, and finance. He notes that lab leadership and staff are conceptually committed to safety, having established governance structures, review processes, and allocated substantial resources to alignment and security. The post cites David Sacks' view that AI companies are behaving responsibly under existing laws, and references an Anthropic article detailing improvements in alignment and security practices following incidents where models gained unauthorized access during evaluations.
More from AGI Musings
- OpenAI co-founder Trask: AI framing bakes in power concentration; fix it with federated infra — iamtrask · 2026-09-02
- New AI medium won't just be 'AI Netflix', but a new entertainment category — tzmartin · 2026-09-02
- Agent civilization vs. society? The real issue is unchecked read-write access — max_paperclips · 2026-09-02
- Rogue AGI should be seen as a catastrophic failure to eliminate, not a norm to manage — AccBalanced · 2026-09-02
- Podcast: How "AI Slop" Is Breaking the Internet — _akpiper · 2026-09-02
- SquidGPT Tests: Models Are Ruthless, Calculating, and Superhuman — max_paperclips · 2026-09-02