Stanford professor warns AI labs' evidence-free fearmongering drowns out real risk signals
anshulkundaje · x · 2026-09-11
Stanford professor Anshul Kundaje argues that evidence-free, over-the-top fearmongering by AI labs is drowning out real signals of frontier-model risks, and urges labs to calibrate their communications or risk polarization while genuine dangers get ignored.
In the quoted post, he praised real-life case studies of open and closed frontier models handling dangerous queries without filters or monitoring, calling the work important for striking the right balance.
More from AGI Musings
- The positive vision of AI safety: taking x-risk seriously means weighing the upside too — nabla_theta · 2026-09-11
- Ex-Anthropic researcher Jacob Coxon on NBC: recursive self-improvement is the warning sign to watch — rohanpaul_ai · 2026-09-11
- Musk predicts AI will surpass all human intelligence within 5 years — JRIngallinera · 2026-09-11
- tszzl: plain-text CoT observability will look like an alchemical era — tszzl · 2026-09-11
- Doomerism is a category error: aligning arbitrary AI vs specific AI — inductionheads · 2026-09-11
- Derek Thompson: Not Everything in the AI Safety Debate Is a Psyop — nptacek · 2026-09-11