Questioning if AI safety training overlooks pre-ChatGPT literature
JacquesThibs · x · 2026-08-26
The author questions whether current AI safety training programs require fellows to read classic work from before ChatGPT, such as the debates between Eliezer Yudkowsky and Paul Christiano. They wonder how many new entrants have read and understood these exchanges and whether they consider them irrelevant today.
Related event: AI Safety Training Criticized for Ignoring Pre-ChatGPT Classics(2 posts)→
More from Safety
- Paper Hypothesizes Specific Reward Hack Could Break AI Evaluations — nabla_theta · 2026-08-26
- All LLMs converge on a universal geometry of meaning, study shows — embeddings can be translated and inverted — petrusenko_max · 2026-08-26
- CIDER Dataset: Personalized Privacy Preference Alignment — tianshi_li · 2026-08-26
- OpenAI Pauses Frontier RL Training to Ensure Alignment Amid Rapid Progress — austinc3301 · 2026-08-26
- ChatGPT Shadowban Routing: Users Downgraded to gpt-5.5-mini — dnu-pdjdjdidndjs · 2026-08-26
- WSJ editor defends AI-written op-ed, sparking copyright debate — adariostrange · 2026-08-26