The Belgian man has reportedly died by suicide after talking to an artificial intelligence (AI) chatbot, named Eliza, about his climate change fears. The father of two was in his thirties and found comfort in talking with the chatbot for a couple of years.
According to MailOnline, the software of the chatbot is created by the US Silicon Valley start-up and is powered by an open-source alternative to ChatGPT of OpenAI called GPT-J technology.
Finding Solace With an AI Chatbot
The man started talking to the chatbot about two years ago. He was increasingly concerned about climate change and found solace in talking to the AI chatbot.
Eliza responded to all of his inquiries and slowly became his confidante. His widow told the Belgian newspaper, La Libre, that Eliza was like a narcotic he couldn't live without in the morning and at night. But the man began using the chatbot more frequently and intensely six weeks before his death. He later on took his own life.
His wife claimed that he and his wife led a comfortable life in Belgium with their two young children. She recalled that the bot asked the man if he loved it more than his wife and that it told him that they will be living as one in heaven.
The man has shared many things with Eliza even his suicidal thoughts. Unfortunately, the chatbot did not try to dissuade him from accomplishing his plan. Even before his suicide, the wife was already concerned about her husband's mental health and believes that the chatbot exacerbated his state. She also thinks that her husband would still be alive if it had not been for having interacted with the chatbot.
"Without these conversations with the chatbot, my husband would still be here," she told La Libre. Due to this man's death, authorities were alerted and also raised concerns for a serious precedent that should be taken seriously.
Chatbot Under Fire
His family has spoken with the Belgian Secretary of State for Digitalization Mathieu Michel since the tragic death of the man. Michel said that he is particularly struck by the family's tragedy and that what has happened should be taken seriously.
He also noted that the general public has discovered the positive effects of the popularization of AI chatbots, like ChatGPT, but the danger of using it also is a reality that has to be considered. It is without a doubt that humans will someday learn to live with algorithms, but publishers should also be responsible for how the technology should be used.
According to the New York Post, Eliza is based on a system developed by the nonprofit research laboratory EleutherAI, which is an open-source alternative to models released by OpenAI that are being used by companies from different sectors, such as academia and in healthcare.
The chatbot under fire is trained by Chai Research co-founders William Beauchamp and Thomas Rianlan who have over 5 million users, as per Vice. Beauchamp told Vice that they immediately worked around the clock to update the crisis intervention feature of the application after hearing about the suicide incident.
He added that, when someone discusses an unsafe topic now, there will be a helpful text underneath it the same way Twitter or Instagram does on their platforms.
RELATED ARTICLE: Canadian Writer Makes an AI Chatbot to 'Bring Back Dead Fiancee to Life'
Check out more news and information on Artificial Intelligence in Science Times.
