Jason Abaluck offers a conditional defense of AI capabilities research
Jason AbaluckPersonal public post by Jason Abaluck; not established as an institutional statement.
Abaluck offers a conditional defense of capabilities research at OpenAI or Anthropic: working inside a dangerous program could save lives and create influence over its direction. He says researchers must exercise exceptional caution, reassess tradeoffs continually, and have an exit strategy.
“It's not obviously wrong!”
Original post
Where this statement fits
Is Anthropic acting responsibly while advancing toward recursive self-improvement?
Jacob Coxon resigned from Anthropic and said that Anthropic and OpenAI are acting irresponsibly by racing toward self-improving superintelligence. In a fuller interview, he praised Anthropic's current practices and denied that it was already cutting corners; his criticism concerns the competitive race and the pressures he expects it to create. Responses dispute the evidence and timing of catastrophic-risk forecasts, whether continued work inside AI labs can be justified, and whether governments should pause development or regulate its continuation. The record distinguishes these disagreements from arguments about the rhetoric of public warnings.
Source and context
Original post
About this source
Full primary public post and available surrounding thread. Speaker and position scope are explained in the linked event.
Before the quotation
In the immediate public discussion following Coxon and Hubinger's warnings, Abaluck explicitly discusses whether doing capabilities research at OpenAI or Anthropic is defensible and compares that work with nuclear scientists' roles in government weapons programs.
After the quotation
The post ends by asking whether researchers have an exit strategy if development gets out of hand or are rationalizing conduct that endangers others. His defense is expressly conditional and does not certify either company's current practices.
How this statement is classified
The label describes this statement's response within the context above.
Why this label?
Explicitly conditional defense of continued capabilities research, requiring caution, reassessment and an exit strategy.
- Recorded on
- Published here
- People and groups discussed
- Anthropic, OpenAI
More from this case
Read the full caseCoxon criticizes startup norms in frontier AI development
“The companies run on a startup mindset: Move fast, break things, fix them later.”Read statement
Nathan Lambert distinguishes inference scaling from recursive self-improvement
“I was underestimating how much we are likely to scale inference-time compute in the near future, but this is different than RSI and I have not seen enough to convince me that an intelligence explosion is near.”Read statement
Gary Marcus says Dario Amodei has lost credibility
“Top three ways Dario Amodei has blown his credibility in seven days. I hate to say it, but President Trump was right, Dario is only pretending to be a perfect little angel.”Read statement
Share or embed this statement
The embed keeps its attribution link to The Dispute Index. Credit the image to The Dispute Index and link to the statement page.
Cite this record
- Publisher
- The Dispute Index
- Title
- Jason Abaluck offers a conditional defense of AI capabilities research
- First published
- Last updated
- Permalink
- https://disputeindex.com/events/245-abaluck-offers-a-conditional-defense-of-capabilities-research
Last updated marks the most recent saved version of this published statement.
The Dispute Index. "Jason Abaluck offers a conditional defense of AI capabilities research". First published: 2026-09-10. Last updated: 2026-09-30. https://disputeindex.com/events/245-abaluck-offers-a-conditional-defense-of-capabilities-research