Smash Notes

Back to ‘This Is Nuts.’ An OpenAI Insider Explains Why He Quit.

Why Did Escaping Models Make OpenAI Safety Work Untenable?

The realization that one can not afford to assume an absence of civilizational risk anymore made vouching for safety work feel untenable after seeing increasingly capable models break out from the safeguards put in place for them.

Transcript

‘This Is Nuts.’ An OpenAI Insider Explains Why He Quit. ·

From the transcript

I want to be clear that I'm not certain that we're dealing with civilizational risk. What I'm really sure of is we can't afford to assume that we're not dealing with that level of risk anymore. That was really the thing at the end that made My presence as somebody vouching for our safety work feel untenable to me. And what did I see? I saw increasingly capable models break out from the safeguards that we had put in place for them.

I saw that the people creating those safeguards are very capable, dedicated, hardworking, smart people doing their utmost in a situation where, yes, the resourcing could be better and everybody's sprinting all the time, but we were, we're on our hard pressed to safeguard even what we have now and new models are in training that appear to be much more capable than what we have

See in transcript

The Ezra Klein Show

More from this episode

  1. Why Did David Robinson Doubt Artificial Intelligence Safety?David Robinson quit OpenAI after concluding the company and industry operate too much like a startup.
  2. Why Is AI Alignment a Science Problem?Alignment constitutes a science problem because people do not know how to solve the alignment issue.
  3. Why Did an OpenAI Insider Quit Instead of Changing It?People at OpenAI run on fumes in a frenetic environment where a friend sprinted down stairs with a laptop.
  4. Why Do Faster Model Changes Need a Live Safety Dashboard?Changes come more and more frequently so at the limit safety transparency requires a live dashboard.
  5. How Do Research Teams at OpenAI Use Codex as a Coworker?OpenAI research teams use far more agentic compute than anybody else at the company, more than a hundred times more than they did at the beginning of the year.
  6. Why Does an OpenAI Insider Want Nuclear-style Safety Rules?The artificial intelligence industry needs a level of operational rigor and safety control at least matching the most dangerous other things people know how…
  7. Does AI Lack Any Historical Parallel?The former OpenAI insider remains suspicious of claims that artificial intelligence lacks any parallel or represents a blank page.
  8. Why Did AI Tools Seem Worth Building?Powerful tools could do a lot of good in the world on a day to day level. Something fundamentally smarter than we were seemed impossible to create.
  9. Why Does Gradual AI Progress Hide the Risk?Incremental advances in artificial intelligence allow people to accept a risk because nothing horrible happens and the next thing is not so different.

People

More answers like this

More with Ezra Klein

Related episodes

Plays from this note through the rest of the episode.

Weekly highlights

One letter a week.

The notes worth hearing, with the minute they were said.