This feels quite close to the definition of the singularity; if an LLM can become both the Generator and the Discriminator (to use a GAN analogy), then we have takeoff.
We might eventually see the code for the models and inference, but I do doubt we’ll see training code or be granted access to training data. Google is pretty bad at this.
That’s my general understanding as well, but it isn’t a large conceptual leap to go from real-time selection of pretrained “z-vectors” to real-time generation of the same. The larger conceptual breakthrough, with demonstration of its effectiveness, is the big success here.
Great research here. Contextual real-time weight modification is definitely one of the breakthroughs required for AGI. Why create a LoRA when you can generate one on the fly suited to the task at hand?
I develop sophisticated LLM programs every day at a small YC startup — extracting insights from thousands of documents a day.
These LLM programs are very different than naive one-shot questions asked of ChatGPT, resembling o1/3 thinking that integrates human domain knowledge to produce great answers that would have been cost-prohibitive for humans to do manually.
Naive use of LLMs by non-technical users is annoying, but is also a straw-man argument against the technology. Smart usage of LLMs in o1/3 style of emulated reasoning unlocks entirely new realms of functionality.
LLMs are analogous to a new programming platform, such as iPhones and VR. New platforms unlock new functionality along with various tradeoffs. We need time to explore what makes sense to build on top of this platform, and what things don’t make sense.
What we shouldn’t do is give blanket approval or disapproval. Like any other technology, we should use the right tool for the job and utilize said tool correctly and effectively.
The technology is incredible, but the path to AGI isn't single-player. Qualia is the missing dataset required for AGI. See attention-schema theory for how social pressures lead to qualia-driven minds capable of true intelligence.
“1 + 1 = 2” is only true in our imagination, according to logical deterministic rules we’ve created. But reality is, at its most fundamental level, probabilistic rather than deterministic.
Luckily, our imaginary reality of precision is close enough to the true reality of probability that it enables us to build things like computer chips (i.e., all of modern civilization). And yet, the nature of physics requires error correction for those chips. This problem becomes more obvious when working at the quantum scale, where quantum error correction remains basically unsolved.
I’m just reframing the problem of finding a grand unified theory of physics that encompasses a seemingly deterministic macro with a seemingly probabilistic micro. I say seemingly, because it seems that macro-mysteries like dark matter will have a more elegant and predictive solution once we understand how micro-probabilities create macro-effects. I suspect that the answer will be that one plus one is usually equal to two, but that under odd circumstances, are not. That’s the kind of math that will unlock new frontiers for hacking the nature of our reality.
I think a better analogy would be vision. Even with a full understanding of the eye and visual cortex, one can only truly understand vision by experiencing sight. If we had to reconstruct sight from scratch, it would be more important to experience sight than to understand the neural structure of sight. It gives us something to aim for.
We basically did that with language and LLMs. Transformers aren’t based on neural structures for language processing. But they do build upon the intuition that the meaning of a sentence consists of the meaning that each word in a sentence has in relation to every other word in a sentence — the attention mechanism. We used our experience of language to construct an architecture.
I think the same is true of qualia and consciousness. We don’t need to know how the hardware works. We just need to know how the software works, and then we can build whatever hardware is necessary to run it. Luckily there’s theories of consciousness out there we can try out, with AST being the best fit I’ve seen so far.
It’s analogous to the automobile. People do still walk, bike, and ride horses, but the vast majority of productive land transportation is done by automobile. Same thing with electronic communication vs. written correspondence. New tech largely supplants old tech. In this case, the old tech is human ingenuity and inventiveness.
I don’t think this is a controversial take. Many people take issue with the premise that artificial intelligence will surpass human intelligence. I’m just pointing out the logical conclusion of that scenario.
It’s true by definition. If we invent a better-than-all-humans inventor, then human invention will give way. It’s a fairly simple idea, and not one I made up.
If the universe is material, then we already know with 10-billion percent certainty that some arrangement of matter causes qualia. All we have to do is figure out what arrangements do that.
Ironically, we understand consciousness perfectly. It is literally the only thing we know — conscious experience. We just don’t know, yet, how to replicate it outside of biological reproduction.
The only reason Sam would leave OpenAI is if he thought AGI could only be achieved elsewhere, or that AGI was impossible without some other breakthrough in another industry (energy, hardware, etc).
High-intelligence AGI is the last human invention — the holy grail of technology. Nothing could be more ambitious, and if we know anything about Altman, it is that his ambition has no ceiling.
Having said all of that, OpenAI appears to be all in on brute-force AGI and swallowing the bitter lesson that vast and efficient compute is all you need. But they’ve overlooking a massive dataset that all known biological intelligences rely upon: qualia. By definition, qualia exist only within conscious minds. Until we train models on qualia, we’ll be stuck with LLMs that are philosophical zombies — incapable of understanding our world — a world that consists only of qualia.
Building software capable of utilizing qualia requires us to put aside the hard problem of consciousness in favor of mechanical/deterministic theories of consciousness like Attention-Schema Theory (AST). Sure, we don’t understand qualia. We might never understand. But that doesn’t mean we can’t replicate.
As a counter to my original thought, perhaps simulating the computation of a skull-sized organic system is possible by building an earth-sized silicon system.
Biology is an organic computation that runs at the speed of existence. There may be no way to simulate this computation in silicon at the same scale, insofar as biology is already massively parallel and may very well involve quantum computation.
Put another way, if we accept for a moment that the universe is a simulation, it may be fundamentally impossible for an in-simulation simulator to ever reach the computational power of its parent simulator.
The article isn’t as interesting as the idea that it is possible to identify someone smarter than oneself.
Reminds me of PG’s blub paradox:
“As long as our hypothetical Blub programmer is looking down the power continuum, he knows he's looking down. Languages less powerful than Blub are obviously less powerful, because they're missing some feature he's used to. But when our hypothetical Blub programmer looks in the other direction, up the power continuum, he doesn't realize he's looking up. What he sees are merely weird languages. He probably considers them about equivalent in power to Blub, but with all this other hairy stuff thrown in as well. Blub is good enough for him, because he thinks in Blub.”
If being smarter is like being tall, then it is easy to identify someone that is smarter/taller than oneself. But if there is some threshold where smartness becomes a difference of kind rather degree, it may not be possible to identify people who are significantly smarter than themselves.
Like the blub programmer, we may mistake people who are smarter than us as… merely weird. They think differently, which we mistake for thinking wrongly.
Add into the mix that there are probably multiple types of “intelligence”, each suited to different domains, and the problem is compounded.
Speed of thought is different than correctness of thought. The first is easy to identify. The second is a problem that I don’t think we should assume is solved.
Counterpoint — good-enough technology prevents over-optimization, where we pay too high of a cost for too-marginal gains. It is generally better to reduce the cost of a 98% solution than to maintain the cost of a 99% solution. There will be exceptions for domains that require a 100% solution, but piano tuning is not one of them.
Incredible work!