AI Risk, Plainly: 10 jargon-free explainers of AI safety ideas, with evidence for and against
birchlse · x · 2026-10-26
Researcher Leonard Dung launched "AI Risk, Plainly," a site of short explainers on AI safety concepts for non-technical readers. Each entry covers a plain definition, why it matters for extreme-risk arguments, and what research has found for and against, with sources. 10 entries are live, including scheming, evaluation awareness, chain-of-thought monitorability, why capable AI might seek power, and the P(doom) debate; reward hacking and stochastic parrots are in preparation.
More from Safety
- Ex-DOGE cyber researchers breach Anthropic systems with AI agents in under 12 hours — ScottNover · 2026-10-26
- Ofcom admits UK online safety laws failed to protect children in first review — gnukeith · 2026-10-26
- Sakana AI CEO Calls for Japan-US Shared Framework for AI Capability Evaluation — SakanaAILabs · 2026-10-26
- Gary Marcus backs call to outlaw voice cloning, citing scam and fraud harms over benefits — GaryMarcus · 2026-10-26
- Turing-complete malware can run in PyTorch's backward pass via a zero-valued loss — retr0jirachi · 2026-10-26
- ISO/IEC JTC 1/SC 42 holds 18th plenary on AI standards in Montreal — davidmanheim · 2026-10-26