311–320 of 328 posts
Earlier quoted context omitted.
Want to collab on a database and some clustering and analysis? I’m a data scientist at FAIR with an interest in antiquarian docs and books
Sadly I'm just an amateur armchair historian (at best) so I doubt I'd be of much help. I'm mostly only doing the translation for my own edification
Earlier quoted context omitted.
> You think dolphins and orcas are not intelligent and don't understand things ? Not sure where you got that non-secitur from ... I would expect most animal intelligence (incl. humans) to be very similar, since their brains are very similar. Orcas are animals. LLMs are not animals.
Orca and human brains are similar, in the sense we have a common ancestor if you look back far enough, but they are still very different and focus on entirely different slices of reality and input than humans will ever do. It's not something you can brush off if you really believe in input supremacy so much. From the orca's perspective, many of the things we say we understand are similarly '2nd hand hearsay'.
To follow your hypothetical, if an Orca were to be exposed to human language, discussing human terrestrial affairs, and were able to at least learn some of the patterns, and maybe predict them, then it should indeed be considered not to have any understanding of what that stream of words meant - I wouldn't even elevate it to '2nd hand hearsay'.
Still, the Orca, unlike an LLM, does at least does have a brain, and does live in and interact with the real world, and could probably be said to "understand" things in it's own watery habitat as well as we do.
Regarding "input supremacy" :
It's not the LLMs "world of words" that really sets it apart from animals/humans, since there are also multi-model LLMs with audio and visual inputs more similar to a humans sensory inputs. The real difference is what they are doing with those inputs. The LLM is just a passive observer, whose training consisted of learning patterns in it's inputs. A human/animal is an active agent, interacting with the world, and thereby causing changes in the input data it is then consuming. The human/animal is learning how to DO things, and gaining understanding of how the word reacts. The LLM is learning how to COPY things.
There are of course many other differences between LLMs/Transformers and animal brains, but even if we were to eliminate all these differences the active vs passive one would still be critical.
> To benchmark LLM handwriting accuracy, last year Dr. Lianne Leddy and I developed a set of 50 documents comprising some 10,000 words—we had to choose them carefully and experiment to ensure that these documents were not already in the LLM training data (full disclosure: we can’t know for sure, but we took every reasonable precaution). You're drawing conclusions from _this_? Let alone pretending that "it did a somet…
Earlier quoted context omitted.
The capabilities of LLMs are unclear but it is clear that they are not just search engines or autocompletes or stochastic parrots. You can disagree. But this is not an opinion. You are factually wrong if you disagree. And by that I mean you don’t know what you’re talking about and you are completely misinformed and lack knowledge. The long term outcome if I’m right is that AI abilities continue to grow and it basical…
> It’s already to the point of where if you’re not vibe coding you’re behind. I like being productive, not babysitting a semi-literate program incapable of learning
Again if you don’t agree then you are lost and uninformed. There are special cases where there are projects where human coding is faster but that is a minority.
Earlier quoted context omitted.
The dream is to be able to teach LLMs to reason more generally, but the reasons this works for math don't generally apply Why is that? Any suggestions for further reading that justifies this point? Ultimately, reinforcement learning is still just a matter of shoveling in more text. Would RL work on humans? Why or why not? How similar is it to what kids are exposed to in school?
An important difference between reinforcement learning (RL) and pre-training is the error feedback that is given. For pre-training the error feedback is just next token prediction error. For RL you need to have a goal in mind (e.g. successfully solving math problems) and the training feedback that is given is the RL "reward" - a measure of how well the model output achieved the goal. With RL used for LLMs, it's the w…
The LLM was explicitly shown how to step by step solve these problems, then trained extensively until it got it
Again, that's all we do. We train extensively until we "get it." Monkey-see, monkey-do turns out not only to be all you need, so to speak... it's all there is.
In contrast, how do you check the answer to more general problems such "Should NATO expand to include Ukraine?"
If you ask a leading-edge model a question like that, you will find that it has become diplomatic enough to remain noncommittal. If this ( https://gemini.google.com/share/9f365513b86f ) isn't adequate, what would you expect a hypothetical genuinely-intelligent-but-not-godlike model to say?
There is only way to check the answer to that question, and that's to sign them up and see how Russia reacts. (Frankly I'd be fine with that, but I can see why others aren't.)
Also see the subthread at https://news.ycombinator.com/item?id=45483938 . I was really impressed by that answer; it wasn't at all what I was expecting. I'm much more impressed by that answer than by the HN posters I was engaging with, let's put it that way.
Ultimately it's not fair to judge AI by asking it for objective answers in questions requiring value judgement. Especially when it's been "aligned" to within an inch of its simulated life to avoid bias. Arguably we are not being given the access we need to really understand what these things are capable of.
Earlier quoted context omitted.
An important difference between reinforcement learning (RL) and pre-training is the error feedback that is given. For pre-training the error feedback is just next token prediction error. For RL you need to have a goal in mind (e.g. successfully solving math problems) and the training feedback that is given is the RL "reward" - a measure of how well the model output achieved the goal. With RL used for LLMs, it's the w…
Going back to the grandparent reply, there's a phrase that carries a LOT of water: The LLM was explicitly shown how to step by step solve these problems, then trained extensively until it got it Again, that's all we do. We train extensively until we "get it." Monkey-see, monkey-do turns out not only to be all you need, so to speak... it's all there is . In contrast, how do you check the answer to more general problem…
They aren’t aligned to avoid bias (which is an incoherent concept, avoiding bias is like not having priors), they are aligned to incorporate the preferred bias of the entity doing it the alignment work.
(That preferred bias may be for a studious neutrality on controversial viewpoints in the surrounding society as perceived by the aligner, but that’s still a bias, not the absence of bias.)
Earlier quoted context omitted.
But these models are more like generalists no? Couldn’t they simply be hooked up to more specialized models and just defer to them the way coding agents now use tools to assist?
There would be no point in going via an LLM then, if I had a specialist model ready I'd just invoke it on the images directly. I don't particularly need or want a chatbot for this.
Earlier quoted context omitted.
An important difference between reinforcement learning (RL) and pre-training is the error feedback that is given. For pre-training the error feedback is just next token prediction error. For RL you need to have a goal in mind (e.g. successfully solving math problems) and the training feedback that is given is the RL "reward" - a measure of how well the model output achieved the goal. With RL used for LLMs, it's the w…
Going back to the grandparent reply, there's a phrase that carries a LOT of water: The LLM was explicitly shown how to step by step solve these problems, then trained extensively until it got it Again, that's all we do. We train extensively until we "get it." Monkey-see, monkey-do turns out not only to be all you need, so to speak... it's all there is . In contrast, how do you check the answer to more general problem…
Which is fine for us humans, but would only be fine for LLMs if they also had continual learning and whatever else was necessary for them to be able to learn on the job and be able to pick up new reasoning skills by themselves, post-deployment.
Obviously right now this isn't the case, so therefore we're stuck with the LLM companies trying to deliver models "out of the box" that have some generally useful reasoning capability that goes beyond whatever happened to be in their pre-training data, and the way they are trying to do that is with RL ...
Earlier quoted context omitted.
Going back to the grandparent reply, there's a phrase that carries a LOT of water: The LLM was explicitly shown how to step by step solve these problems, then trained extensively until it got it Again, that's all we do. We train extensively until we "get it." Monkey-see, monkey-do turns out not only to be all you need, so to speak... it's all there is . In contrast, how do you check the answer to more general problem…
> Again, that's all we do. We train extensively until we "get it." Monkey-see, monkey-do turns out not only to be all you need, so to speak... it's all there is. Which is fine for us humans, but would only be fine for LLMs if they also had continual learning and whatever else was necessary for them to be able to learn on the job and be able to pick up new reasoning skills by themselves, post-deployment. Obviously rig…
It'll obviously happen at some point. No reason why it won't.
Just as obviously, current LLMs are capable of legitimate intelligent reasoning now, subject to the above constraints. The burden of proof lies on those who still claim otherwise against all apparent evidence. Better definitions of 'intelligence' and 'reasoning' would be a necessary first step, because our current ones have decisively been met.
Someone who has lost the ability to form memories is still human and can still reason, after all.
Earlier quoted context omitted.
> Again, that's all we do. We train extensively until we "get it." Monkey-see, monkey-do turns out not only to be all you need, so to speak... it's all there is. Which is fine for us humans, but would only be fine for LLMs if they also had continual learning and whatever else was necessary for them to be able to learn on the job and be able to pick up new reasoning skills by themselves, post-deployment. Obviously rig…
Agreed, memory consolidation and object permanence are necessary milestones that haven't been met yet. Those are the big showstoppers that keep current-generation LLMs from serving as a foundation for something that might be called AGI. It'll obviously happen at some point. No reason why it won't. Just as obviously, current LLMs are capable of legitimate intelligent reasoning now , subject to the above constraints. T…
Continual learning, resulting in my AI being different from yours, because we've both got them doing different things, is also likely to turn the current training and deployment paradigm on it's head.
I agree we'll get there one day, but I expect we'll spend the next decade exploiting LLMs before there is any serious effort more on to new architectures.
In the meantime, DeepMind for one have indicated they will try to build their version of "AGI" with an LLM as a component of it, but it remains to be seen exactly what they end up building and how much new capability that buys. In the long term building in language as a component, rather than building in the ability to learn language, and everything else that humans are capable of learning, is going to prove a limitation, and personally I wouldn't call it AGI until we do get to that level of being able to learn everything that a human can.