Mixed or conditional

We need coordination to be able to approach future capability increases with an appropriate degree of caution and humility, and we need it yesterday.

Responding to Hubinger's Coxon-related warning, Wolfe calls the pace of AI development terrifying and urges urgent coordination between companies and countries. He also praises costly recent actions by OpenAI and does not supply an extinction probability.

Speaker
Jason Wolfe
Capacity
Member of Technical Staff at OpenAI working on alignment and model behavior, according to his MATS mentor biography; personal account, no claimed institutional authority in the post.
Observed
September 10, 2026
Published here
September 10, 2026
Targets
Anthropic, OpenAI
Classification rationale
Supports urgent coordination while crediting OpenAI's recent precautions.

Surrounding material

Full context

Before the quotation

Wolfe says he does not know his probability for literal extinction but sees multiple ways AI could go poorly. The included quoted post is Hubinger's acknowledgement of risk and partial defense of Anthropic.

After the quotation

Wolfe says he is heartened by costly actions OpenAI has recently taken, but argues that no single company or country can solve the problem alone.

Evidence

Complete source record

Jason Wolfe on Coxon’s warning and the frontier AI safety debate (opens in a new tab) · Jason Wolfe

Full primary public post and available surrounding thread. Related context: https://www.matsprogram.org/mentor/wolfe https://x.com/EvanHub/status/2097497037956891126 Speaker and position scope are explained in the linked event.