Hacker News Reader

Comment by ninetyninenine | Hacker News Reader

Fiction and story writing is hallucination. It is the opposite of a stochastic parrot.

We've achieved both extremes of AI. Computers can be both logical machines and hallucinators. Our goal is to create a machine that can be both at the same time and can differentiate between both.

That's the key. Hallucination is important but the key is for the computer to be self aware about when it's hallucinating.

Of course it's a hard problem but even humans hallucinate massively. Just look at religion. Only one religion can be right or none, that must mean,logically speaking all other religions are hallucinations.

View on HN

Comparing religion to LLM mistakes is a case of the very prevalent anthropomorphism in society currently. I fear this misunderstanding and conflation will prevent us actually improving the tech.

Coming to an incorrect explanation such as, “Helios pulls the Sun across the sky every day,” is categorically different than a math program incorrectly returning the most likely next token in a sequence. LLMs don’t have beliefs at all.

Helios answers a question “Why does the Sun rise?” Holding such a belief shows a logical understanding that some force must be doing this, but due to a lack of knowledge of the world, the person comes up with an incorrect explanation.

LLMs can’t pose and reason about such questions. It is just not the same class of “hallucinations.” Assuming we’ve figured out cognition via word prediction is going to get us nowhere fast in the long term.

No. I never said we figured out cognition.

The LLM is still a black box feed forward network. It is the intricacies of how signals interact with each other in this network that we don't fully understand.

Word prediction and curve fitting are high level concepts we used to build an intelligence we don't fully understand.

Also belief and understanding are orthogonal concepts. I can believe something I don't understand and I can understand something I don't belief.

My claim here is that LLMs understand things from a certain aspect because LLMs can produce output indistinguishable from understanding.

Also because both the human brain and the LLM are black boxes there is no other metric we can use to gauge the level of understanding an LLM has other than comparing it's inputs and outputs to the human brain.

It is not the opposite of stochastic parrot, it is exactly the same thing only the predictions are worse due to sparse training data.

> Only one religion can be right or none, that must mean,logically speaking all other religions are hallucinations.

There are some mistakes in this sentence.

It is possible (if unlikely) that multiple religions accurately describe some aspects of the world, while being mistaken about others. That is, treating rigorous complete "correctness" as the only useful state a religion could have is very misleading. Newtonian physics and special relativity both fail to predict some observed phenomena, but they're still both useful (and not every religion claims rigorous perfect correctness, even if some do).

Even if some religions can be shown to be wrong, that doesn't automatically mean that they're hallucinations. People can believe things for plausible reasons and be wrong about them.

People can also have reasonable stances like "I cannot prove this is true, and would not try to, but my subjective personal experience of visions of God persuade me it's probably real."

That seems very different to me from an LLM hallucinating a paper from whole cloth out of the blue.

> Just look at religion.

This is bit off-topic but what I see as one of driving force behind existence of religions is need for personification. It seems easier for human to interact with the world and its elements by communicating with it as it was familiar parson-like entity.

Now when we talk about LLMs and AI in general, there is often personification as well.

LLMs are trained to actually imitate human understanding deliberately. The data is human and the high level training is defined as the most likely word prediction of human output data.

So not surprising to find aspects of personification in the LLM. It is trained on US.

> even humans hallucinate massively

Simpler example: Dreams.

Yeah good point. But dreams are easily distinguishable from reality.

Religion is often indistinguishable from truth and reality to those who hallucinate it.

Confusing sincere but incorrect belief with hallucination is categorically wrong.

Most humans are extremely aware of the difference between dreams and reality. If LLMs had similar awareness when they hallucinated there wouldn't be a problem.

Reply on news.ycombinator.com