Researchers at Stanford University recently tested out some of the more popular AI tools on the market, from companies like OpenAI and Character.ai, and tested how they did at simulating therapy.The researchers found that when they imitated someone who had suicidal intentions, these tools were more than unhelpful — they failed to notice they were helping that person plan their own death.
“[AI] systems are being used as companions, thought-partners, confidants, coaches, and therapists,” says Nicholas Haber, an assistant professor at the Stanford Graduate School of Education and senior author of the new study. “These aren’t niche uses – this is happening at scale.”
AI is becoming more and more ingrained in people’s lives and is being deployed in scientific research in areas as wide-ranging as cancer and climate change. There is also some debate that it could cause the end of humanity.
As this technology continues to be adopted for different purposes, a major question that remains is how it will begin to affect the human mind. People regularly interacting with AI is such a new phenomena that there has not been enough time for scientists to thoroughly study how it might be affecting human psychology. Psychology experts, however, have many concerns about its potential impact.
One concerning instance of how this is playing out can be seen on the popula …
