Jasmine Sun anthropologist of disruption ✰ atlantic contributor ✰ san francisco

“Experts over 80% of the time made the same decision as the AI, and over 70% of the time used the AI's arguments to justify their decision. But only 15% of the time did they say, oh yeah, I was influenced by this AI agent. … There's all this concern about AI takeover and rogue AIs. And I think that a lot of the times we're just going to let the AIs take over because we'll just be outsourcing all our decisions and we won't even realize.”

Hard Fork (NYT) · Tech Strategy & Big Tech · October 2026

Context

Asked what she was tracking this week, Sun describes a randomized controlled trial from Anthropic in which professional analysts were given AI advice on a decision. She summarizes the study's reported findings, then gives her own worry.

Listen to the episode → More from Jasmine Sun → More Tech Strategy & Big Tech quotes →

Transcript

Hard Fork (NYT) Around 48:46 into the episode

Erin Griffith

self-employed. What are you talking about? You are your own boss.

Speaker 2

Well, I am also self-hating. So I think I can see you. Jasmine, what are you tracking this week?

Speaker 3

Oh, man. I saw that this new research report that a team at Anthropic put out. They did a randomized controlled trial trying to understand how much experts defer to AI systems. So they basically gave a bunch of professional, like statistical analysts access to AIs to give them advice on a decision in a scenario. And what they found was that not only did the experts, I think over 80% of the time, make the same decision as the AI, which are split half half, they also over 70% of the time use the AI's arguments to justify their decision. But only 15% of the time did they say, oh, yeah, I was influenced by this AI agent. And so we have this world, which I think is one that we're going to be in increasingly, where humans are oftentimes almost becoming like rappers on AI agents and AI systems, where like all we're doing is like seeing the faces and like typing the thoughts. And like, I feel like what, like, there's all this concern about AI takeover and rogue AIs. And I think that a lot of the times we're just going to let the AIs take over because we'll just be outsourcing all our decisions and we won't even realize. So that's what I'm worried about.

Speaker 2

That is a cheery thought. I mean, as somebody who professionally takes other people's opinions and repeats them as though they were my own, like I do, I do kind of sympathize. It's just usually the people I'm stealing from are people, not, you know, not artificial.

Speaker 3

People say that like AI makes everybody a manager. Like here we are.

Speaker 2

Yeah, exactly. That's exactly. This is exactly what managers do. Wow. Aaron, what about you? What are you tracking?

Speaker names come from our own diarization. The position is estimated from where the line sits in the episode.

More from Hard Fork (NYT)

“An $80 billion net loss seems insane, but it's actually kind of a weird accounting quirk where they have to count some convertible debt that they issued to some of their shareholders as a net loss. … What we've actually seen is that in the first half of this year, they had an operating loss of around six to $700 million. And that's shrinking from the same period last year where it was over $2 billion.” — Erin Griffith, Hard Fork (NYT)

“Toward the end, there were a lot of spectators showing up, including people who were in a book club who read John's book and who came together as a little book club outing. And I'm asking them, what are you guys doing here? And one of them told me, which I'll never forget, oh, we're just fans of white collar crime… In the same way that there are people who are fans of murder stories or murder podcasts, there's also a fraud contingent as well.” — Erin Griffith, Hard Fork (NYT)