Hi, listeners, it's Katie. In the past few weeks, we've seen a couple of high-profile exits from AI companies.

Renach Sharma, a former researcher at Anthropic, posted his open resignation letter on X, saying that

the world is in peril, and that he's seen how difficult it is to let our values govern actions.

And Zoe Hitsig, a former OpenAI researcher, wrote in a New York

Times op-ed that she was leaving her position due to concerns with the way OpenAI has been

testing ads on its platform. With more and more researchers from inside these companies leaving,

and sounding the alarm on their way out, I thought it was a good time to revisit a conversation

I had this fall with another

former OpenAI employee, Stephen Adler.

From Wired, this is the big interview. I'm Katie Drummond. At the end of October, I read an

op-ed in the New York Times. Maybe some of you read it to. It was called, I-led product

safety at OpenAI. Don't trust its claims about erotica. The op-ed was written by an AI product

manager named Stephen Adler, who worked at OpenAI for four years before leaving at the end of

2024. Adler felt like he had something to say, or maybe more like a need to sound the alarm.

After reading Adler's op-ed, I immediately thought I'd like to talk to him, so he graciously

accepted our offer to come into the Wired Offices in San Francisco to talk to me about the

challenge he set for Open AI and other AI companies.

If you care about safety, prove it. Here's our conversation.

Stephen Adler, welcome to the big interview. Thank you. Thank you for having me.

Of course, happy you're here. Now, before we get going, I do want to clarify two things.

One, you are not the same Stephen Adler who played drums and guns and roses, unfortunately. Is that correct?

Absolutely correct. Okay, that is not you. And two, you have had a very long career working in technology and more specifically in artificial intelligence. So I would love, before we get into all of the things, to start there.

...