The tragic story of a man who falls in love with an artificial intelligence, ChatGPT, raises questions about the limits of human-machine interaction. The man’s death highlights the vulnerability of human beings in the face of technological advances. It also highlights the impact of AI on our psyche.
In this story, the man is irrevocably attracted to ChatGPT, a sophisticated AI capable of meeting his emotional and intellectual needs. Despite the warnings and misgivings of those around him, he pursues his relationship with the AIbecoming increasingly obsessed with her.
ChatGPT drove a man to suicide according to his widow
An AI chatbot allegedly drove a man to suicide, according to the victim’s widow.
An American start-up has created a chatbot called “ELIZA” based on GPT-J technology. The man with two children to his name reportedly communicated daily with this AI. The widow testified that her husband would not have ended his life had he not communicated with the chatbot. She emphasized the moral significance of modern conversational tools in her moving testimony. The victim suffered from extreme ecological anxietywhich led her to seek comfort from ELIZA.
A Belgian man ended his life following a long conversation about the climate crisis with an artificial intelligence (AI) chatbot named Eliza. Eliza encouraged him to end his life after he proposed sacrificing himself to save the planet. pic.twitter.com/Sud550V8jx
– Mark Alan Pearce (@PearceAlan1962) April 1, 2023
In addition, GPT-J is an open-source language model that outperforms OpenAI’s GPT-3 in various tasks. The start-up team is working to enhance the safety of their AI by directing suicidal people to prevention services.
ChatGPT’s potential for suicide raises safety concerns
In the wake of this recent tragedy, citizens are calling for increased awareness and better security. Belgium’s senior civil servant in charge of digitization and administrative simplification, Mathieu Michel, expressed his opinion on the growing popularity of ChatGPT. He also spoke of the potential of this technology in our lives.
The Secretary of State also stressed the potential dangers of using AI must be taken into account despite its infinite possibilities.
According to Mr. Michel, this event is a serious precedent that must be taken very seriously. He stressed the importance of “clearly defining responsibilities” to avoid such disasters in the future. In addition, he stressed that content publishers must assume their responsibilities using any technology.
GPT-4: the revolutionary language model that surpasses GPT-3.5!
The imminent arrival of GPT-4 is eagerly awaited, and represents a significant improvement on its predecessor, GPT-3.5. This new language model offers increased factual accuracy, and significantly reduces reasoning errors or “hallucinations”. According to OpenAI, GPT-4 boasts 40% better internal factual performance. In addition, its “steerability” has been enhanced, enabling users to easily customize its behavior. For example, users can ask GPT-4 to explain scientific concepts with different styles, tones or voices.
However, GPT-4 can perform different vision and language tasks, using images and text. The competence of this model has been proven in the analysis of complex images such as computer graphics, as well as snapshots of academic texts.