When German journalist Martin Bernklautyped his name and location into Microsoft’s Copilot to see how his articles would be picked up by the chatbot, the answers horrified him. Copilot’s results asserted that Bernklau was an escapee from a psychiatric institution, a convicted child abuser, and a conman preying on widowers. For years, Bernklau had served as a courts reporter and the AI chatbot had falsely blamed him for the crimes whose trials he had covered.
The accusations against Bernklau weren’t true, of course, and are examples of generative AI’s “hallucinations.” These are inaccurate or nonsensical responses to a prompt provided by the user, and they’re alarmingly common. Anyone attempting to use AI should always proceed with great caution, because information from such systems needs validation and verification by humans before it can be trusted.
But why did Copilot hallucinate these terrible and false accusations?
I don’t disagree, I was just pointing out that “each word is generated independently of each other” isn’t strictly accurate for LLM’s.
It’s part of the reason they are so convincing to some people, they are able to hold threads semi-coherently throughout entire essay length paragraphs without obvious internal lapses of logic.
I think you’re seeing coherence where there is none.
Ask it to solve the riddle about the fox the chicken and the grains.
Even if it does solve the riddle without blurting out random nonsense, that’s just because the sample data solved the riddle billions of times before.
It’s just guessing words.