Tech journalist Gil Duran on "LLM-poisoned individuals" - especially on LinkedIn

[highplainsdem]

One thing AI LLMs are very good at: convincing their victims that mediocrity is greatness, and that the ability to generate high volumes of mediocre content is a substitute for skill, talent, or work. This is the distinct impression I get every time I look at LinkedIn.

— Gil Durán (@gilduran.com) 2026-05-24T18:50:13.963Z

There seems to be a popular genre on LinkedIn in which LLM-poisoned individuals use LLMs to write gleaming odes to LLMs. It's a bit frightening. Also, I know more than one person now who is in the throes of some form of AI psychosis.

— Gil Durán (@gilduran.com) 2026-05-24T18:52:49.030Z

By this, I mean:
1. They spend much of their day talking to chatbots.
2. They generate reams of mediocre content that they feel are very important and must be shared.
3. They perceive themselves as being on cusp on a major breakthrough.
4. They don't understand why you don't get it.

— Gil Durán (@gilduran.com) 2026-05-24T18:54:34.468Z

5. It never occurs to them what might happen if they told the chatbot to provide the opposite advice or poke holes in their theory. So, at the same time as they are living in the machine, they appear to fundamentally misunderstand it.

— Gil Durán (@gilduran.com) 2026-05-24T18:56:57.034Z

These are not the people being talked into suicide. But they are educated individuals living in a parallel reality. And while they increasingly seem unhinged, they believe those of us who are not drowning in a chatbot relationship are missing the boat. In this way, it resembles a religious delusion.

— Gil Durán (@gilduran.com) 2026-05-24T19:27:28.993Z

They are the chosen ones who will be saved and enjoy the fruits of the coming kingdom, where they shall live with their digital God. The rest of us are the lost souls who did not heed the gospel and will be left out from AI abundance paradise.

— Gil Durán (@gilduran.com) 2026-05-24T19:30:24.907Z