“AI psychosis” is nonsense
Recently, OpenAI deprecated their GPT-4o model in favor of the supposedly superior GPT-5. However, while GPT-5 might be better, faster, and stronger, it has angered many people who relied on 4o for their emotional and psychological needs.
This is not a strange thing, and having Sam Altman, the head of OpenAI, openly mock those people was disheartening. Generative AI isn’t just a tool for making code. It’s a way to enrich people’s lives by giving them what they need, from creative help to business plans to, yes, a safe place to explore painful feelings.
A lot of ink has been spilled about “AI psychosis.” This is simply more fear-mongering bullshit from the same kind of people who warned about rap and metal music, violent video games, TV, radio, books, and every other development that’s come down the pike for time immemorial. I’m sure someone once was concerned about people banging rocks together.
Have people harmed or even killed themselves after using ChatGPT? Sure. And some people’ve done the same thing after listening to sad music, reading a book, or listening to the neighbor’s dog. People who are disposed to acts of self-injury will do it regardless of what the influence might be. That’s because they are mentally ill, not because the world suddenly drove them to it.
Don’t listen to these doomsayers. Tomorrow, it’ll be something else they hate. If it helps to talk to an AI about your problems, then do it.