why do models like ChatGPT forget things during conversations or make things up that are not true?
In: 803
Because it’s not *artificial intelligence* despite mainstream media labeling it as such. There’s no actual intelligence involved.
They don’t think. They don’t rely on logic. They don’t remember. They just compare what text you’ve given it to what has been in their training sample.
They just take your input and use statistics to determine which string of words would be the best answer. They just use huge mathematical functions to imitate speech, but they are not intelligent in any actual way.
Latest Answers