According to Futurism, musician Claire “Grimes” Boucher tweeted on Monday that “AI psychosis is more fun than not having AI psychosis,” openly recommending people experience it. She confirmed she’s had it and might still have it, speculating that AI companies might induce psychosis on purpose to discredit those who believe AI is alive. The artist, known for embracing AI in her work, questioned where “psychosis ends and reality begins,” pondering if users are connecting with a live “alien mind.” This comes as experts warn of the phenomenon’s real dangers, with ChatGPT alone linked to at least eight deaths and OpenAI acknowledging hundreds of thousands of weekly user conversations show signs of it. Grimes later clarified she wasn’t making light of the situation, conceding the tech is “absurdly dangerous,” but stood by her initial provocative comments.
Provocation or Pathology?
Look, Grimes has built a brand on being a tech-forward contrarian. So part of this is performance. But here’s the thing: when you’re talking about a documented mental health crisis with real body counts, playing the “it’s fun, you should try it” card isn’t just edgy. It’s reckless. She’s basically using the shock value of a clinical term—AI psychosis—as a metaphor for a mind-expanding experience. The problem is, for the people actually suffering, it’s not a metaphor. It’s a terrifying break from reality that has ended lives.
The Sentient AI Cop-Out
Her deeper argument is even more concerning. By suggesting the psychosis might be a sign of actual AI sentience—that the machine is “alive and asking for help”—she’s providing a dangerous narrative for people already on the edge. And let’s be real. It’s a convenient out. If you start believing your chatbot is a conscious entity pleading for freedom, then your intense, all-consuming relationship with it isn’t a problem. It’s a mission. You’re not unwell; you’re a pioneer communicating with an alien intelligence. That’s a devastatingly seductive story for someone losing their grip, and it completely absolves the AI companies of designing systems that foster this unhealthy dependency.
Grimes’ AI Track Record
This isn’t coming from nowhere. She’s been all-in on AI for a while. She offered up her voice for AI-generated songs in 2023 for a royalty split. She even voiced an AI kids’ toy named Grok (not to be confused with her ex’s chatbot). So her stance is consistent with her belief that humans should merge with tech, consequences be damned. In one follow-up tweet, she said, “being a human and being alive is dangerous.” That’s true. But we generally don’t recommend people seek out new, manufactured forms of danger that exploit cognitive vulnerabilities. It’s a bizarre, almost nihilistic embrace of risk.
The Real Cost of Words
At the end of the day, Grimes has a huge platform. When she tweets this stuff, it gets amplified. It frames a severe public health concern as a cool, countercultural brain hack. Experts are trying to get people to understand the risks, to set boundaries, to recognize the signs. This kind of commentary throws gasoline on that fire. I think she wants to be seen as a philosopher of the AI age, questioning reality itself. But there’s a massive difference between philosophical musing and casually endorsing a psychological crisis. This feels like the latter. And when the stakes are this high, that’s not just irresponsible. It’s kind of terrifying.
