Hacker News new | ask | show | jobs
by sporadicallyjoe 1056 days ago
Is anyone shipping AI products that DO NOT contain hallucinations? I thought that was pretty much a given.
2 comments

Well there isn't a human that never "hallucinates" in meaning we use for LLMs aka gives "incorrect answers" confidently.

Human's brains use lots of heuristics - we don't "think step by step" through everything - instead we rapidly construct an answer for almost everything.

What we say is "hallucinations" for AI in humans is "misspeaking, misremembering anything, off by 1 math/counting, missidentifying someone, using the wrong variable/method when programming, etc."

Hallucinating is roughly how they work, we just label it as such when it's something obviously weird
This is something I'm not sure people understand.

LLM's only make a "best guess" for each next token. That's it. When it's wrong we call it a "hallucination" but really the entire thing was a "hallucination" to begin with.

This is also analogous to humans - who also "hallucinate" incorrect answers, usually "hallucinate" incorrect answers less when they "Think through this step by step before giving your answer", etc.