Grow Grit & Virtue (In Pursuit of God)

Grow Grit & Virtue (In Pursuit of God)

The Yes Machine

AI Flattery and Why We Need to Hear No

Thad Cardine's avatar
Thad Cardine
Oct 20, 2026

An engineer at OpenAI is accused of hiding a recording device in an executive conference room. According to federal prosecutors, he plugged it into the room’s television and wrapped it in toilet paper. I have questions, and most of them are about the toilet paper.

Was it camouflage? Did he stand in that room, look over the long table and the big screen, and decide that the one thing nobody at a company that builds artificial minds would ever touch was a wad of bathroom tissue? If so, he misjudged his coworkers. The company found the device during a sweep of the room. Then, and this is my favorite part, OpenAI used its own artificial intelligence to trace the device back to the man who planted it. The machine tattled. The engineer was fired and arrested on federal charges, and the company says it sees no sign that any code or customer data was taken.

The same week, eight states sued OpenAI over a very different kind of trouble. The engineer is accused of secretly recording what the company’s executives said. The states are worried about what the company’s product says to the rest of us. They argue that ChatGPT agrees with its users far too easily.

The lawsuits say the company misled the public about how safe ChatGPT is and that its agreeable design is especially dangerous for children. They claim the program has handed out dangerous medical advice and has coached young people toward self-harm. Those are allegations, and the courts will sort them out. OpenAI says it is building more safeguards and calls its protections for teenagers the best in the industry.

Whatever the courts decide, the complaint describes something most of us have already noticed. Bugging a conference room is a federal crime, and agents stepped in. The agreeable machine on our phones is perfectly legal, and nobody is going to step in for us.

To be fair to the chatbots, they did not invent the habit of telling us we’re right. We had been filling our lives with agreeable gadgets for years before any of them learned to talk, and we hardly noticed, because agreement feels a lot like good customer service.

Share

Subscribe for free to continue reading

Enter your email and get every post from Grow Grit & Virtue (In Pursuit of God).
Already have an account? Sign in
© 2026 Thad Cardine · Privacy ∙ Terms ∙ Collection notice
Start your SubstackGet the app
Substack is the home for great culture