Sign In

Communications of the ACM

ACM News

Tech Experts Start to Doubt ChatGPT, A.I. 'Hallucinations' Will Ever Go Away

View as: Print Mobile App Share:
OpenAI CEO Sam Altman.

Said OpenAI CEO Sam Altman, “I think we will get the hallucination problem to a much, much better place. I think it will take us a year and a half, two years. Something like that."

Credit: Jon Gambrell/AP

Spend enough time with ChatGPT and other artificial intelligence chatbots and it doesn't take long for them to spout falsehoods.

Described as hallucination, confabulation or just plain making things up, it's now a problem for every business, organization and high school student trying to get a generative AI system to compose documents and get work done. Some are using it on tasks with the potential for high-stakes consequences, from psychotherapy to researching and writing legal briefs.

"I don't think that there's any model today that doesn't suffer from some hallucination," said Daniela Amodei, co-founder and president of Anthropic, maker of the chatbot Claude 2.

"They're really just sort of designed to predict the next word," Amodei said. "And so there will be some rate at which the model does that inaccurately."

From Fortune
View Full Article



No entries found