Post #2719617

2026-05-20 18:07 UTC

It’s like a comedy routine. “Question and no guessing or mistakes gemini” “Confident answer that’s totally made up” “Were you guessing or hallucinating in the last statement?” “I apologize, you’re 100% correct that I was guessing” “How do I get you to stop guessing” "Use this code…" Repeat endlessly I do think it comes up with some buried, interesting sources, but that’s about it. It’s like a 5 year old.

Replies (2)

@affenlehrer@feddit.org 2026-05-20 20:28

It can’t not hallucinate. It’s just predicting (not even selecting) next tokens. It doesn’t know what it knows and what it doesn’t know. It can’t introspect. It just gives probabilities for all possible tokens in it’s vocabulary based on the context window and the inference engine selects the next one (based on it’s settings). Without having the correct answer in the context window it can just make a prediction based on it’s (fixed) neutral net parameters and these are severely limited, even for big models. What I mean is, they basically “learned” the whole Internet and compressed the whole thing into some hundred billion or a few trillion parameters. That’s an insane compression ratio. This compression is lossy. For niece information and the results are similar to the “unimportant” details in highly compressed JPGs, you can make out the general image but fine details are just a mush. The LLM itself doesn’t know this, it just gives wrong predictions. For what it does I think the result is extremely impressive but the way it works is severely limited.

Open ##2722662
@MonkderVierte@lemmy.zip 2026-05-20 19:34

It’s a text generator… that’s like telling a parrot to not repeat.

Open ##2725087