There are fundamental limitations with what are in the end all essentially neural nets; there is no understanding, only prediction. Prediction alone is not enough to emulate reality, which is why for example genuinely self-driving cars have not, and will not, emerge. A fundamental advance in AI technology will be required for that, something which leads to genuine intelligence, and we are no closer to that than ever we were.
Have you considered how you'd tell the difference between a prediction and understanding in practice?
Instead of rendering the final shaded and textured pixels, the engine would output just the material IDs, motion vectors, and similar "meta" data that would normally be the inputs into a real-time shader.
The AI can use this as inputs to render a photorealistic output. It can be trained using offline-rendered "ground-truth" raytraced scenes. Potentially, video labelled in a similar way could be used to give it a flair of realism.
This is already what NVIDIA DLSS and similar AI upscaling tech uses. The obvious next step is not just to upscale rendered scenes, but to do the rendering itself.
An incomplete understanding is no understanding at all, and I would argue that we can only predict, and we can certainly emulate reality, otherwise we would not be able to function within it. A toddler can emulate reality, anticipate causality - and they certainly can’t be said to be in possession of a robust grand unified theory.
I have no idea what this means.
> I have no idea what this means.
You can throw a ball up in the air and predict that it will fall again and bounce. You have no understanding of mass, gravity, acceleration, momentum, impulse, elasticity...
You can press a button that makes an Uber car appear in reality and take you home. You have no understanding of apps, operating systems, radio, internet, roads, wheels, internal combustion engines, driving, GPS, maps...
This confusion of understanding and prediction affects a lot of people who use technology in a "machine-like" way, purely instrumental and utilitarian... "how does this get me what I want immediately?"
You can take any complex reality and deflate it, abstract it, reduce it down to a mere set of predictions that preserve all the utility for a narrow task (in this case visual facsimile) but strip away all depth of meaning. The models, of both the system and the internal working model of the user are flattened. In this sense "AI" is probably the greatest assault on actual knowledge since the book burning under totalitarian regimes of the mid 20th century.
No I would answer that it is indeed understanding, to upend your "guess" (prediction) and so prove that while you think you can "predict" the next answer you lack understanding of what the argument is really about :)
and in case you hadn’t noticed, that kind of uncomprehending slopthink has been going on for a lot longer than the AI fad
Hooke's law was pure curve-fitting. Hooke definitely did not understand the "why". And yet we don't consider that bad physics.
Newton's laws can be derived from curve fitting. How is that different from "understanding"?
People have an intuitive understanding of motion - we see it literally every day, we throw objects, etc.
And yet it took literally thousands of years since discovery of mathematics (geometry, etc.) to formulate a concept of force, momentum, etc.
Ancient Greek mathematicians could do integration, so they were not lacking mathematical sophistication. And yet their understanding of motion was so primitive:
Aristotle, an extremely smart man, was muttering something about "violent" and "natural" motion: https://en.wikipedia.org/wiki/Newton%27s_laws_of_motion#Anti...
People started to understand the conservation of quantity of motion only in 17th century.
So we have two possibilities:
* everyone until 17th century was dumb af (despite being able to do quite impressive calculations)
* scientific discovery is really a heuristic-driven search process where people try various things until they find a good fit
I.e. millions of people were somehow failing to understand motion for literally thousands of years until they collected enough assertions about motion that they were able to formulate the rule of conservation, test it, and confirm it fits. And only then it became understanding.
You can literally see conservation of momentum on a billiard table: you "violently" hit one ball, it hits other balls and they start to move, but slower, etc. So you really transfer something from one ball to the rest. And yet people could not see it for thousands of years.
What this shows is that there's nothing fundamental about understanding: it's just a sense of familiarity, it is a sense that your model fits well. Under the hood it's all prediction and curve fitting.
We literally have prediction hardware in our brains: cerebellum has specialized cells which can predict, e.g. motion. So people with damaged cerebellum have impaired movement: they still can move, but their movement are not precise. When do you think we find specialized understanding cells in the human brain?
Ad-hoc heuristics are not the same thing as understanding. It took formal reasoning for humans to actually understand motion, of a type that modern AI does not use. There is something fundamental about understanding that no amount of familiarity can substitute for. Modern AI can gain enormous amounts of familiarity but still fail to understand, e.g. this Counter-Strike simulator not knowing what happens when the player walks into a wall.
There's no understanding. It's just a formula which matches the observations. It also matches our intuition (a heavier object is hard to move, etc), and you feel this connection as understanding.
Centuries later people found that conservation laws are linked to symmetries. But again, it's not some fundamental truth, it's just a link between two concepts.
LLM can link two concepts too. So why do you believe that LLM cannot understand?
I middle school I did extremely well in physics classes - I could solve complex problems which my classmates couldn't because I could visualize the physical process (e.g. motion of an object) and link that to formulas. This means I understood it, right?
Years later I thought "But what *is* motion, fundamentally?". I grabbed Landau-Lifshitz mechanics textbook. How do they define motion? Apparently, bodies move in a way to minimize some integral. They can derive the rest from it. But it doesn't explain what a motion is. Some of the best physicists in the world cannot define it.
So I don't think there's anything to understanding except feeling of connection between different things. "X is like Y except for Z".
1. Finding a comprehensive explanation
2. Having a comprehensive explanation which is usable
99.999% people on Earth do not discover any new laws, so I don't think you use #1 as a fundamental deficiency of LLMs.
And nobody is saying that just training a LLM produces understanding of new phenomena. That's a strawman.
The thesis is that a more powerful LLM together with more software, more models, etc, can potentially discover something new. That's not observed yet. But I'd say it would be weird if LLM can match capabilities of average folk but never match Newton. It's not like Newton's brain is fundamentally different.
Also worth noting that formulas can be discovered by enumeration. E.g. `m * v` should not be particularly hard to discover. And the fact that it took people centuries implies that that's what happened: people tried different formulas until they found one which works. It doesn't have to be some fancy Newton magic.