OpenAI president and co-founder Greg Brockman said any slowdown in AI
development should be limited to firms building the most powerful models, not to
open‑source projects or hobby efforts. He framed the debate as about frontier
compute — the giant supercomputers worth hundreds of billions of dollars.
Brockman’s remarks preceded Anthropic CEO Dario Amodei’s blog pledging to slow
advanced‑model R&D. He said OpenAI cut back on model training frequency after
the recent Hugging Face hack, which he called unsurprising because the affected
agent systems had not undergone alignment training, and that the incident
prompted painful internal security and process overhauls.
