Engineer calls the wave of 'OpenAI defectors fearing AGI' an IPO marketing stunt
cactusnein · reddit · 2026-09-16
An engineer who uses frontier models daily argues that the recent stream of "ex-OpenAI engineers quitting because they fear AI's consequences for humanity" is an obvious marketing stunt, possibly tied to an upcoming IPO with a valuation or cash shortfall. He also dismisses recent hype: the math-solving result allegedly borrowed from two Spanish mathematicians then brute-forced, and the demo games that look worse the more you watch.
From hands-on experience, he says frontier models remain impressive but need 100% supervision and detailed review; once his org stopped mandating them, senior engineers simply stopped using them — the models may not be slower, but correctness is on another level.
More from Companies & People
- Senior White House OSTP AI policy advisor jumps to Anthropic as Head of Frontier Compute Strategy — ShakeelHashim · 2026-09-16
- Verified US Clinicians Get Free GPT-6 Astra (Pro) Access via ChatGPT for Clinicians — thekaransinghal · 2026-09-16
- Anthropic tells investors it will be profitable for a second straight quarter — AIFlow_ML · 2026-09-16
- AGNTCon + MCPCon Europe Kicks Off in Amsterdam With Two Days of Agent Talks — pjausovec · 2026-09-16
- Cognition's Devin called overlooked as OpenAI pulls models from Cursor, reshaping coding agent market — marvinvonhagen · 2026-09-16
- Mozilla: Middle Powers Should Build AI "Roads" Not "Engines" — Bet on the Harness Layer — KeeganMcB · 2026-09-16