Anthropic's overlooked April 2022 RLHF paper foreshadowed its coding-first comeback and IPO
xennygrimmato_ · x · 2026-11-20
- The author revisits Anthropic's April 2022 paper "Training a Helpful and Harmless Assistant with RLHF" — the first time he saw a productized chatbot, English/Python proficiency, and a helpful-but-harmless training goal laid out together in a paper.
- The work never went viral: ChatGPT stole the thunder in Dec 2022 and built a huge first-mover advantage.
- He credits Anthropic's turnaround to its focus on coding, a strong model, and a great product — culminating in an IPO ahead of OpenAI. The attached image shows the paper's crowdworker preference-selection interface.
Related event: Revisiting Anthropic's 2022 RLHF Paper(2 posts)→
More from Companies & People
- Anthropic lists on Nasdaq under ticker ANT as MarketSite marks debut — edwinarbus · 2026-11-20
- Ollama CEO: Enterprise customers now ask where their data goes, not where models run — ollama · 2026-11-20
- 25 exec calls fed to Claude reveal the 10 real reasons enterprise AI transformation stalls — vasuman · 2026-11-20
- Kevin Roose interviewed 150 insiders for his new book The AGI Chronicles — every · 2026-11-20
- Anthropic's NYC pop-up gave away caps and coffee, drawing Apple-scale lines — signulll · 2026-11-20
- Engineer digs up his 2024 Anthropic rejection email as the company goes public — tdhopper · 2026-11-20