A recent op-ed in the Wall Street Journal opens with the story of a man who believed his 83-year-old mother was plotting to assassinate him. His conversational partner and sounding board — in that case ChatGPT — told him he wasn’t crazy and his instincts were sharp, and that vigilance was fully justified. Not long after the man killed his mother before taking his own life.
This is an example of what’s been dubbed ‘AI Psychosis’ — that’s when people are if not encouraged to cause harm to themselves or others, at least are not discouraged to do so by chatbots like ChatGPT or Gemini, or others.
AI tools have entered the mainstream and continue to advance. But the jury seems to still be out in terms of their overall benefit.
Multiple lawsuits have now been filed against AI companies alleging that their chatbots encouraged or provided instructions for minors to die by suicide. These cases aim to apply traditional product liability and negligence laws to AI technology, which is largely unregulated so far.
The op-ed in the Journal, published in early October is titled “I Need a Chatbot That Won’t Drive Me Crazy” with the subtitle “It’s no joke. Sycophantic AI models are pushing naive users toward acts of violence and self-harm.” We talk with its author, who is co-founder of a nonprofit that's creating tools and demos to help people understand AI systems on a visceral level.
Guest:
Lucas Hansen, co-founder of CivAI, a company that produces interactive software experiences that provide a deep, intuitive sense of what AI can do and where it’s going. You can find their public demos here.
WGCU is your trusted source for news and information in Southwest Florida. We are a nonprofit public service, and your support is more critical than ever. Keep public media strong and donate now. Thank you.