Preferences

In my case I fully grasp what such a future could be, but I don't think we are on the path to that, I believe people are too optimistic, i.e. they just believe instead of being truly skeptical.

From where I look at it, LLMs are flawed in many ways, and people who see progress as inevitable do not have a mental model of the foundation of those systems to be able to extrapolate. Also, people do not know any other forms of AI or have though hard about this stuff on their own.

The most problematic things are:

1) LLMs are probabilistic and a continuous function, forced by gradient descent. (Just having a "temperature" seems so crazy to me.) We need to merge symbolic and discrete forms of AI. Hallucinations are the elephant in the room. They should not be put under the rug. They should just not be there in the first place! If we try to cover them with a layer of varnish, the cost will be very large in the long run (it already is: step-by-step reasoning, mixture of experts, RAG, etc. are all varnish, in my opinion)

2) Even if generalization seems ok, I think it is still really far from where it should be, since humans need exponentially less data and generalize to concepts way more abstract than AI systems. This is related to HASA and ISA relations. Current AI systems do not have any of that. Hierarchy is supposed to be the depth of the network, but it is a guess at best.

3) We are just putting layer upon layer of complexity instead of simplifying. It is the victory of the complexifiers and it is motivated by the rush to win the race. However, I am not so sure that, even if the goal seems so close now, we are going to reach it. What are we gonna do? Keep adding another order of magnitude of compute on top of the last one to move forward? That's the bubble that I see. I think that that is not solving AI at all. And I'm almost sure that a much better way of doing AI is possible, but we have fallen into a bad attractor just because Ilya was very determined.

We need new models, way simpler, symbolic and continuous at the same time (i.e. symbolic that simulate continuous), non-gradient descent learning (just store stuff like a database), HAS-A hierarchies to attend to different levels of structure, IS-A taxonomies as a way to generalize deeply, etc, etc, etc.

Even if we make progress by brute forcing it with resources, there is so much work to simplify and find new ideas that I still don't understand why people are so optimistic.


Symbolic AI is dead. Either stop trying to dig out and reanimate its corpse, or move the goalposts like Gary Marcus did - and start saying "LLMs with a Python interpreter beat LLMs without, and Python is symbolic, so symbolic AI won, GG".

Hallucinations are incredibly fucking overrated as a problem. They are a consequence of the LLM in question not having a good enough internal model of its own knowledge, which is downstream from how they're trained. Plenty of things could be done to improve on that - and there is no fundamental limitation that would prevent LLMs from matching human hallucination rates - which are significantly above zero.

There is a lot of "transformer LLMs are flawed" going around, and a lot of alternative architectures being proposed, or even trained and demonstrated. But so far? There's nothing that would actually outperform transformer LLMs at their strengths. Most alternatives are sidegrades at best.

For how "naive" transformer LLMs seem, they sure set a high bar.

Saying "I know better" is quite easy. Backing that up is really hard.

> Hallucinations are incredibly fucking overrated as a problem. They are a consequence of the LLM in question not having a good enough internal model of its own knowledge, which is downstream from how they're trained. Plenty of things could be done to improve on that - and there is no fundamental limitation that would prevent LLMs from matching human hallucination rates - which are significantly above zero.

Why is there no fundamental limitation that would prevent LLMs from matching human hallucination rates? I'd like to hear more about how you arrived at that conclusion.

To avoid hallucinations, you, a human, need two things: you need to have an internal model of your own knowledge, and you need to act on it - if your meta-knowledge says "you are out of your depth", you either answer "I don't know" or look for better sources before formulating an answer.

This is not something that's impossible for an LLM to do. There is no fundamental issue there. It is, however, very easy for an LLM to fail at it.

Humans get their (imperfect, mind) meta-knowledge "for free" - they learn it as they learn the knowledge itself. LLM pre-training doesn't give them much of that, although it does give them some. Better training can give LLMs a better understanding of what the limits of their knowledge are.

The second part is acting on that meta-knowledge. You can encourage a human to act outside his knowledge - dismiss his "out of your depth" and provide his best answer anyway. The resulting answers would be plausible-sounding but often wrong - "hallucinations".

For an LLM, that's an unfortunate behavioral default. Many LLMs can recognize their own uncertainty sometimes, flawed as their meta-knowledge is - but not act on it. You can run "anti-hallucintion training" to make them more eager to act on it. Conversely, careless training for performance can encourage hallucinations instead (see: o3).

Here's a primer on the hallucination problem, by OpenAI. It doesn't say anything groundbreaking, but it does sum up what's well known in the industry: https://openai.com/index/why-language-models-hallucinate/

I read the page you linked, but I'm still not understanding why hallucination isn't an inevitability of LLMs. The explanation OpenAI gives doesn't feel like a complete answer to me.

OpenAI claims that hallucination isn't an inevitability because you can train a model to "abstain" rather than "guess" when giving an "answer". But what does that look like in practice?

My understanding is that an LLM's purpose is to predict the next token in a list of tokens. To prevent hallucination, does that mean it is assigning a certainty rating to the very next token it's predicting? How can a model know if its final answer will be correct if it doesn't know what the tokens that come after the current one are going to be?

Or is the idea to have the LLM generate its entire output, assign a certainty score to that, and then generate a new output saying "I don't know" if the certainty score isn't high enough?

The answer is that it does know. Not exactly, but the "general shape" of the answer is known to the LLM before the very first token of the answer is emitted!

"Next token prediction" is often overstated - "pick the next token" is the exposed tip of a very large computational process.

And LLMs are very sharp at squeezing the context for every single bit of information available in it. Much less so at using it in the ways you want them to.

There's enough information at "no token emitted yet" for an LLM to start steering the output towards "here's the answer" or "I don't know the answer" or "I need to look up more information to give the answer" immediately. And if it fails to steer it right away? An LLM optimized for hallucination avoidance could still go "fuck consistency drive" and take a sharp pivot towards "no, I'm wrong" mid-sentence if it had to. For example, if you took control and forced a wrong answer by tampering with the tokens directly, then handed the control back to the LLM.

By "shape" of the answer, what do you mean? I always visualized token prediction as a vector pointing off into some sort of cloud of related tokens, and if that's a fair way to visualize it, I could understand how you could say, before even emitting the first token of the answer, "we are pointing towards generally the correct place where the answer is found". But when a single token can make or break an answer, I still don't see how you can truly know whether the answer is correct until the very last token is reached. Because of this, I'm still not convinced hallucination can be stopped.

Can you help correct where I'm going wrong?

Symbolic AI isn't dead, we use it all the time, it's just not a good orchestrating layer for interacting with humans. LLMs are great as a human interface and orchestrator but they're definitely going to be calling out to symbolic models for expanded functionality. This pattern is obvious, we're already on the path with agentic tool use and toolformers.
This is what I mean by "move the goalposts like Gary Marcus did", yes.

If what you're claiming is that external, vaguely-symbolic tooling allows a non-symbolic AI to perform better on certain tasks, then I agree with that.

If you replace "a non-symbolic AI" with "a human", I agree with that too.

symbols and concepts are just collections of neurons that fire with the correct activation. its all about the bitter lesson, human beings cannot design ai, they can only find the most general equations, most general loss function, and push data in. and thats what we have, and thats why its a big deal. The LLM is just a manifestation of a much broader discovery, a generalized learning algorithm. it worked on language because of the information density, but with more compute, we may be able to push in more general sensory data...
Not sure this is a good counterpoint in defence of LLMs, but I'm reminded of how Unix people explain why (in their experience) data should be encoded, stored and transmitted as text instead of something more seemingly natural like binary. It's because text provides more ways to read and transform it, IN SPITE of its obvious inefficiency. LLMs are the ultimate Unix text transformation filter. They are extremely flexible out-of-the-box, and friendly towards experimentation.
> We are just putting layer upon layer of complexity instead of simplifying.

It really irks me that the direction every player seems to be going to is to layer LLMs on top of each other with the goal of saving money on inference while still making the users believe that they are returning high quality results.

Instead of discovering some radical new ways of improving the algorithms they are only marginally improving existing architectures and even that is debatable.

Symbolic AI is mostly dead, we spend a lot of time and money on it and got complex and fragile systems that are far worse than LLMs.

This item has no comments currently.

Keyboard Shortcuts

Story Lists

j
Next story
k
Previous story
Shift+j
Last story
Shift+k
First story
o Enter
Go to story URL
c
Go to comments
u
Go to author

Navigation

Shift+t
Go to top stories
Shift+n
Go to new stories
Shift+b
Go to best stories
Shift+a
Go to Ask HN
Shift+s
Go to Show HN

Miscellaneous

?
Show this modal