Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

An LLM is a lossy encyclopedia

simonwillison.net

Post reply on HN
Print view View original

111–120 of 365 posts

Re: An LLM is a lossy encyclopedia

#111

Post by withinboredom »

jbstack wrote: post #78
kgeist wrote: post #68

I think an LLM can be used as a kind of lossy encyclopedia, but equating it directly to one isn't entirely accurate. The human mind is also, in a sense, a lossy encyclopedia. I prefer to think of LLMs as lossy predictors. If you think about it, natural "intelligence" itself can be understood as another type of predictor: you build a world model to anticipate what will happen next so you can plan your actions accordin…

Are LLMs really lossier than humans? I think it depends on the context. Given any particular example, LLMs might hallucinate more and a human might do a better job at accuracy. But overall LLMs will remember far more things than a human. Ask a human to reproduce what they read in a book last year and there's a good chance you'll get either absolutely nothing or just a vague idea of what the book was about - in this c…

That's not exactly true. Every time you start a new conversation; you get a new LLM for all intents. Asking an LLM about an unrelated topic towards the end of a ~500 page conversation will get you vastly different results than at the beginning. If we could get to multi-thousand page contexts, it would probably be less accurate than a human, tbh.

Re: An LLM is a lossy encyclopedia

#112

Post by checkyoursudo »

simonw wrote: post #34
latexr wrote: post #30

A lossy encyclopaedia should be missing information and be obvious about it, not making it up without your knowledge and changing the answer every time . When you have a lossy piece of media, such as a compressed sound or image file, you can always see the resemblance to the original and note the degradation as it happens. You never have a clear JPEG of a lamp, compress it, and get a clear image of the Milky Way, the…

I think you are missing the point of the analogy: a lossy encyclopedia is obviously a bad idea, because encyclopedias are meant to be reliable places to look up facts.

I am sympathetic to your analogy. I think it works well enough.

But it falls a bit short in that encyclopedias, lossy or not, shouldn't affirmatively contain false information. The way I would picture a lossy encyclopedia is that it can misdirect by omission, but it would not change A to ¬A.

Maybe a truthy-roulette enclyclopedia?

Re: An LLM is a lossy encyclopedia

#113

Post by withinboredom »

layer8 wrote: post #108
GuB-42 wrote: post #33

There are a lot of parallels between AI and compression. In fact the best compression algorithms and LLMs have in common that they work by predicting the next word. Compression algorithms take an extra step called entropy coding to encode the difference between the prediction and the actual data efficiently, and the better the prediction, the better the compression ratio. What makes a LLM "lossy" is that you don't ha…

One difference is that compression gives you one and only one thing when decompressing. Decompression isn't a function taking arbitrary additional input and producing potentially arbitrary, nondeterministic output based on it. We would have very different conversations if LLMs were things that merely exploded into a singular lossy-expanded version of Wikipedia, but where looking at the article for any topic X would g…

LLMs deliberately insert randomness. If you run a model locally (or sometimes via API), you can turn that off and get the same response for the same input every time.

Re: An LLM is a lossy encyclopedia

#114

Post by qwertox »

It could be non-lossy if it would actually reach out to an encyclopedia.

If one takes it as a language engine which translates human language into API calls, and API call results to human language, it would appear to be a non-lossy encyclopedia.

It is the basic building block which enables computers to handle natural language.

The simulated intelligence is proof of its capability as a language model, but it is often so dumb that it is doesn't feel like a "knowledge model".

Re: An LLM is a lossy encyclopedia

#115

Post by ijk »

ACCount37 wrote: post #63
simonw wrote: post #29

Earlier quoted context omitted.

I don't think that's universally true with the new models - I've seen Claude 4 and GPT-5 ask for clarification on questions with obvious gaps. With GPT-5 I sometimes see it spot a question that needs clarifying in its thinking trace, then pick the most likely answer, then spit out an answer later that says "assuming you meant X ..." - I've even had it provide an answer in two sections for each branch of a clear ambig…

A lot of the touted "fundamental limitations of LLMs" are less "fundamental" and more "you're training them wrong". So there are improvements version to version - from both increases in raw model capabilities and better training methods being used.

I'm frustrated by the number of times I encounter people assuming that the current model behavior is inevitable. There's been hundreds of billions of dollars spent on training LLMs to do specific things. What exactly they've been trained on matters; they could have been trained to do something else.

Interacting with a base model versus an instruction tuned model will quickly show you the difference between the innate language faculties and the post-trained behavior.

Re: An LLM is a lossy encyclopedia

#116

Post by ijk »

thunderbong wrote: post #100

What I tell my non-technical friends when they ask about AI - The goal of an LLM is not to give you 100% accurate answers. The goal of an LLM is to continue the conversation.

Exactly: at their core the only task they ate capable of is "complete this document."

It just turned out that document completion was far more effective than anyone anticipated.

Re: An LLM is a lossy encyclopedia

#117

Post by Lerc »

latexr wrote: post #30

A lossy encyclopaedia should be missing information and be obvious about it, not making it up without your knowledge and changing the answer every time . When you have a lossy piece of media, such as a compressed sound or image file, you can always see the resemblance to the original and note the degradation as it happens. You never have a clear JPEG of a lamp, compress it, and get a clear image of the Milky Way, the…

The argument is that a banana is a squishy hammer.

You're saying hammers shouldn't be squishy.

Simon is saying don't use a banana as a hammer.

Re: An LLM is a lossy encyclopedia

#119

Post by ACCount37 »

jpcompartir wrote: post #93
gjm11 wrote: post #88

Earlier quoted context omitted.

Lossy compression does make things up. We call them compression artefacts. In compressed audio these can be things like clicks and boings and echoes and pre-echoes. In compressed images they can be ripply effects near edges, banding in smoothly varying regions, but there are also things like https://www.dkriesel.com/en/blog/2013/0802_xerox-workcentres... where one digit is replaced with a nice clean version of a diff…

Interesting, in the LLM case these compression artefacts then get fed into the generating process of the next token, hence the errors compound.

Not really. The whole "inference errors will always compound" idea was popular in GPT-3.5 days, and it seems like a lot of people just never updated their knowledge since.

It was quickly discovered that LLMs are capable of re-checking their own solutions if prompted - and, with the right prompts, are capable of spotting and correcting their own errors at a significantly-greater-than-chance rate. They just don't do it unprompted.

Eventually, it was found that reasoning RLVR consistently gets LLMs to check themselves and backtrack. It was also confirmed that this latent "error detection and correction" capability is present even at base model level, but is almost never exposed - not in base models and not in non-reasoning instruct-tuned LLMs.

The hypothesis I subscribe to is that any LLM has a strong "character self-consistency drive". This makes it reluctant to say "wait, no, maybe I was wrong just now", even if latent awareness of "past reasoning look sketchy as fuck" is already present within the LLM. Reasoning RLVR encourages going against that drive and utilizing those latent error-correction capabilities.

Re: An LLM is a lossy encyclopedia

#120

Post by Pamar »

DrScientist wrote: post #118

I have another analogy. LLM are animatronic rubber ducks. https://en.wikipedia.org/wiki/Rubber_duck_debugging ( and obviously like all analogies - this one is lossy )

I definitely agree with that, at least this is how I use chatGPT in 99% of the cases.
Post reply on HN