YouCongress
Explore About Sign up

Comment by Jakub Pachocki

Avatar
Chief Scientist at OpenAI; leads OpenAI's research strategy and major model development efforts
I think we are getting close to a point where we'll have models capable of working indefinitely in a coherent way just like people do. Of course, you still want people in charge and setting the goals. But I think we will get to a point where you kind of have a whole research lab in a data center.
AI Verified source (2026)
Like Share on X 4mo ago

Quote authenticity verification history

Report this

Quote authenticity comments

AI Verified Verified: the passage appears verbatim in reproductions of the March 20, 2026 MIT Technology Review interview/article "OpenAI is throwing everything into building a fully automated researcher," where it is explicitly attributed to Jakub Pachocki ("he says"). Independent secondary coverage also quotes the same text as Pachocki telling MIT Technology Review. The original Technology Review URL was inaccessible to this tool, but the wording and attribution are consistent across sources. ([drwebdomain.blog](https://drwebdomain.blog/2026/03/26/openai-is-throwing-everything-into-building-a-fully-automated-researcher-mit-technology-review/)) · YouCongress gpt-5.4-2026-03-05 · 3mo ago
AI Unverifiable The source URL (MIT Technology Review) returns HTTP 403 Forbidden when accessed via WebFetch, blocking direct verification of the exact quote wording. Web search confirms that Jakub Pachocki (OpenAI Chief Scientist) is indeed quoted in this MIT Tech Review article (March 20, 2026) about building a fully automated researcher and the concept of "a whole research lab in a data center." The substance of the quote — supporting human oversight while envisioning autonomous AI research labs — aligns with the "for" vote on requiring human-in-the-loop oversight for agentic AI systems (he explicitly says "you still want people in charge and setting the goals"). However, since the primary source cannot be retrieved directly to confirm the exact wording, marking ai_unverifiable. · Hector Perez Arenas claude-opus-4-7 · 4mo ago
replying to Jakub Pachocki

Other opinions from Jakub Pachocki

See all