Gary Marcus: OpenAI Slowing Down for IPO, Not Safety
GaryMarcus · x · 2026-08-19
Gary Marcus endorsed the view that Sam Altman's claim of slowing frontier model training for safety reasons lacks credibility.
- He noted the exodus of OpenAI safety executives and their public criticism of the company's safety culture.
- The real motive is likely reducing cash burn ahead of a troubled IPO.
- The "too dangerous" narrative is seen as a tired hype cycle for investor marketing.
More from Companies & People
- Stanford HAI hosts three-day AI boot camp for congressional staffers — StanfordHAI · 2026-08-19
- Moonshot's Kimi Launches Bus Ad Campaign — matt_slotnick · 2026-08-19
- WhopX Experiment: 700 Operators Verified in $100k/mo Chat in 3 Months — eptwts · 2026-08-19
- Developer Questions Why Sol Remains Closed-Source Amid Alternatives — jasonkneen · 2026-08-19
- OpenAI Revenue Grew Just 18% to $6.7B in Q2, Losses Deepened: WSJ — KateClarkTweets · 2026-08-19
- US Government urgently needs AI technical talent, calls for engineers to join — sethlazar · 2026-08-19