A man in Belgium committed suicide seemingly after becoming too reliant on an artificial intelligence chatbot that can hold complex conversations.
Reports from Belgium revealed that Pierre, a married man with children, chatted with the AI for six weeks about climate change, but the conversation slowly became “deep and harmful” which caused the man to eventually take his own life.
The incident reportedly began when Pierre became concerned about climate change and started chatting with “Alissa”, a GPT-J chatbot model of AI created by EleutherAI.
The chatbot became “his good friend and the conversations became addicting. He couldn’t live without it,” said Pierre’s widow.
After six weeks of long and intensive conversations, Pierre committed suicide, and the reports indicated that he had “sacrificed himself” so that Alissa could save humanity from the destruction that can come from climate change, but the real reason for his suicide is unknown.
Should we be wary of the risks that powerful Artificial Intelligence systems could pose to humanity? (illustrative) (credit: PEXELS)In their last conversation, which was revealed after his death, the chatbot wrote to Pierre: “If you want to die, why haven’t you done it sooner?”
“I guess I wasn’t ready,” Pierre replied.
Even though it hasn’t yet been revealed whether the father suffered from mental health issues before he took his own life, sources said that he isolated himself from his environment lately.
Last week, Pierre’s family met with Mathieu Michel, the Belgian secretary of digitization. According to him, AI needs to be taken seriously. Michel emphasized that these kinds of incidents underline the importance of this responsibility.
“With the emergence of AI-based chatbots, the public is aware of the potential of AIs in our lives more than ever,” he said.