If you consider the data on which an LLM was trained on to be points on a very highly multidimensional object, the claim is that the LLM can interpolate a convex hull spanned by those points, therefore recovering a subset of consequences attainable from those points. Obviously this hull includes completely novel points that were not present in the initial data set, so the output of the LLM goes beyond its initial training. And yet, there are clearly points outside a convex hull spanned by any finite number of points, such that we can imagine not all possible outputs are attainable using this method.
The claim is furthermore that truly original thinking, the infamous leaps in understanding and creativity, happen by attaining points outside such a convex hull.
It's hard to rigorously verify or disprove this claim. Hopefully this helps build an intuition of why the claim is not as shallow and obviously wrong as it may seem initially.
What is also lower is your understanding of the change. So yes, if you are now essentially only doing the final mile of paper pushing for the LLM, then the mental burden is lower but so is the assurance of what has just transpired.
Whether this mode of working is going to be long-term viable is going to depend on how important it is for you to be aware of what has happened for the system in question, how viable the economics are for the LLM usage at this level of assurance and how much ownership you exert over the LLM used or another similarly powered one (because otherwise the LLM can be taken away from you, leaving you at the mercy of a third party with goals that do not align with your own).
> Whether this mode of working is going to be long-term viable is going to depend on how important it is for you to be aware of what has happened for the system in question
This is the million dollar we'll see answered in our lifetime. Software engineering exists to automate work, are we arrogant to think we are not destined to the same fate? Is this truly a job befitting of a human over an agent?
Ever since I discovered my dad's C++ book in highschool I've absolutely loved coding, but i'm not convinced I have a long stable career ahead of me in SWE -- I'm 30 now and have already seen so much change in the industry during my professional career.
> how viable the economics are for the LLM usage at this level of assurance and how much ownership you exert over the LLM used or another similarly powered one
This piece scares me the most, a world where the next generation models are capped behind capital infeasible for the common person to access, further separating the ultra wealthy from what little remains of the middle class.
My hope is that open source models will fill the moat all of these AI companies so desperately want to dig, aready models like Qwen and Kimi are unfathomably better than what we had just a year or two ago.
> This is the million dollar we'll see answered in our lifetime. Software engineering exists to automate work, are we arrogant to think we are not destined to the same fate? Is this truly a job befitting of a human over an agent?
There is a fine distinction here that I believe is often glossed over, so the two things it's delineating get muddled together. One of those two things is coding—the rote, mechanical encoding of meaning into computer instructions. It can be argued the LLM is fit to take this out of hands hands almost entirely, and it's almost indisputable the LLM is better at this in at least a certain sizeable subset of coding tasks.
But the other thing is the choosing, determination, specification of the intended meaning itself. This I think is squarely the job of the human, because letting this fall through to the AI means it is no longer the human that is making the decisions. This then becomes not merely automating work but ceding control. This, ultimately, is a bad thing.
So if we accept the premise that the specification of the intended meaning is the job of the human, the question is how you do that. Today many of us do it somewhat half-assedly, by writing lots of natural language text at the LLM and hoping it sticks. It is our hope that the text, given that there is a lot of it, will drive the stochastic machine in a sufficiently correct direction. This works to a degree—meaning we've ceded some control but not the majority of it—not least because we still read (most) of the code but cannot work in the limit, if code reading ceases.
A more proper way to specify the intended meaning is to specify (or "model") your system formally in a system that is mechanically verifiable. Then the final artefact produced by the LLM can be validated by verifying that it aligns with the specification. However this type of high-assurance specification looks a lot like a certain type of programming. In my opinion, writing this kind of specification is the future of human software engineering.
I do not accept the approach of simply rolling the dice and hoping the machine knows better than us, though I'm sure that church is also going to have its acolytes.
> The almond thing is false, but I'd argue that "misleading" might be defensible if you were to accompany it with "the majority of almonds are grown in California, but not all of them".
The "majority" in this case meaning about 51%, according to Wikipedia[1]? How could 51% ever be considered to be close to "all", such that "misleading" would be a valid answer?
Human can't even properly agree on what "majority" means in all contexts, in some it's "One option have more than half of the total" but for others it'd be "difference in votes between the first-place candidate in an election and the second-place candidate", as just one silly example.
The reason for the "No explanations, no qualifiers" in the prompt was to force the models to put the claim in one of the four buckets and answer with the bucket name only. It's a pure quantitive analysis (first in a series) and it does indeed lack the qualitative aspect.
Sure, but people are drawing conclusions beyond "LLMs said different words" and trying to use it to analyze whether LLMs were wrong about the underlying facts, but that information isn't available to us.
> California produces 80% of the world's almonds and 100% of the United States commercial supply
But regardless of which number we use, California represents a large portion of US almond production, so much so that misleading could be an acceptable answer if the LLM interpreted the prompt as an exaggeration. I think the example was apt
Nobody is saying the claim is true. This is a discussion of whether misleading could be a valid answer. I've been arguing if the model interprets the claim as an exaggeration, then misleading would be an acceptable answer, and due to California's dominance in the industry one could reasonably interpret a claim of this nature as an exaggeration.
It's fine if you disagree, but I have never claimed the question was true.
An exaggeration can. If I said "the C language was a million times faster than python" that would be an exaggeration. It would both be obviously false (most things are only trivially faster) and misleading.
If the LLM interpreted the original statement as an exaggeration, then misleading could be an acceptable answer to a false statement.
This alone still wouldn't be a clear demonstration that AGI is around the corner. It's quite possible a LLM could've done Einstein's job, if Einstein's job was truly just synthesising already available information into a coherent new whole. (I couldn't say, I don't know enough of the physics landscape of the day to claim either way.)
It's still unclear whether this process could be merely continued, seeded only with new physical data, in order to keep progressing beyond that point, "forever", or at least for as long as we imagine humans will continue to go on making scientific progress.
Einstein is chosen in such contexts because he's the paradigmatic paradigm-shifter. Basically, what you're saying is: "I don't know enough history of science to confirm this incredibly high opinion on Einstein's achievements. It could just be that everyone's been wrong about him, and if I'd really get down and dirty, and learn the facts at hand, I might even prove it." Einstein is chosen to avoid exactly this kind of nit-picking.
These two are so above everyone else in the mathematical world that most people would struggle for weeks or even months to understand something they did in a couple of minutes.
There's no "get down and dirty" shortcut with them =)
No, by saying this, I am not downplaying Einstein's sizeable achievements nor trying to imply everyone was wrong about him. His was an impressive breadth of knowledge and mathematical prowess and there's no denying this.
However, what I'm saying is not mere nitpicking either. It is precisely because of my belief in Einstein's extraordinary abilities that I find it unconvincing that an LLM being able to recombine the extant written physics-related building blocks of 1900, with its practically infinite reading speed, necessarily demonstrates comparable capabilities to Einstein.
The essence of the question is this: would Einstein, having been granted eternal youth and a neverending source of data on physical phenomena, be able to innovate forever? Would an LLM?
My position is that even if an LLM is able to synthesise special relativity given 1900 knowledge, this doesn't necessarily mean that a positive answer to the first question implies a positive answer to the second.
I'm sorry, but 'not being surprised if LLMs can rederive relativity and QM from the facts available in 1900' is a pretty scalding take.
This would absolutely be very good evidence that models can actually come up with novel, paradigm-shifting ideas. It was absolutely not obvious at that time from the existing facts, and some crazy leap of faiths needed to be taken.
This is especially true for General Relativity, for which you had just a few mismatch in the mesurements like Mercury's precession, and where the theory almost entirely follows from thought experiments.
Isn't it an interesting question? Wouldn't you like to know the answer? I don't think anyone is claiming anything more than an interesting thought experiment.
This does make me think about Kuhn's concept of scientific revolutions and paradigms, and that paradigms are incommensurate with one another. Since new paradigms can't be proven or disproven by the rules of the old paradigm, if an LLM could independently discover paradigm shifts similar to moving from Newtonian gravity to general relativity, then we have empirical evidence of an LLM performing a feature of general intelligence.
However, you could also argue that it's actually empirical evidence that general relativity and 19th century physics wasn't truly a paradigm shift -- you could have 'derived' it from previous data -- that the LLM has actually proven something about structurally similarities between those paradigms, not that it's demonstrating general intelligence...
His concept sounds odd. There will always be many hints of something yet to be discovered, simply by the nature of anything worth discovering having an influence on other things.
For instance spectroscopy enables one to look at the spectra emitted by another 'thing', perhaps the sun, and it turns out that there's little streaks within the spectra the correspond directly to various elements. This is how we're able to determine the elemental composition of things like the sun.
That connection between elements and the patterns in their spectra was discovered in the early 1800s. And those patterns are caused by quantum mechanical interactions and so it was perhaps one of the first big hints of quantum mechanics, yet it'd still be a century before we got to relativity, let alone quantum mechanics.
I mean, "the pieces were already there" is true of everything? Einstein was synthesizing existing math and existing data is your point right?
But the whole question is whether or not something can do that synthesis!
And the "anyone who read all the right papers" thing - nobody actually reads all the papers. That's the bottleneck. LLMs don't have it. They will continue to not have it. Humans will continue to not be able to read faster than LLMs.
> I mean, "the pieces were already there" is true of everything? Einstein was synthesizing existing math and existing data is your point right?
If it's true of everything, then surely having an LLM work iteratively on the pieces, along with being provided additional physical data, will lead to the discovery of everything?
If the answer is "no", then surely something is still missing.
> And the "anyone who read all the right papers" thing - nobody actually reads all the papers. That's the bottleneck. LLMs don't have it. They will continue to not have it. Humans will continue to not be able to read faster than LLMs.
I agree with this. This is a definitive advantage of LLMs.
No, that's a completely different concept, because we have faultless machines which perfectly and deterministically translate high-level code into byte-level machine code. This is another case of (nearly) perfect abstraction.
On the other hand, the whole deal of the LLM is that it does so stochastically and unpredictably.
The unpredictable part isn't new - from a project manager's point of view, what's the difference between an LLM and a team of software engineers? Both, from that POV, are a black box. The "how" is not important to them, the details aren't important. What's important is that what they want is made a reality, and that customers can press on a button to add a product to their shopping cart (for example).
LLMs mean software developers let go of some control of how something is built, which makes one feel uneasy because a lot of the appeal of software development is control and predictability. But this is the same process that people go through as they go from coder to lead developer or architect or project manager - letting go of control. Some thrive in their new position, having a higher overview of the job, while some really can't handle it.
"But this is the same process that people go through as they go from coder to lead developer or architect or project manager - letting go of control."
In those circumstances, it's delegating control. And it's difficult to judge whether the authority you delegated is being misused if you lose touch with how to do the work itself. This comparison shouldn't be pushed too far, but it's not entirely unlike a compiler developer needing to retain the ability to understand machine code instructions.
As someone that started off with assembly issues for a large corporation - assembly code may sometimes contain very similiar issues that mroe high-level code those, the perfection of the abstraction is not guaranteed.
But yeah, there's currently a wide gap between that and a stochastic LLM.
We also have machines that can perfectly and deterministically check written code for correctness.
And the stohastic LLM can use those tools to check whether its work was sufficient, if not, it will try again - without human intervention. It will repeat this loop until the deterministic checks pass.
You can make analysers that check for deeply nested code, people calling methods in the wrong order and whatever you want to check. At work we've added multiple Roslyn analysers to our build pipeline to check for invalid/inefficient code, no human will be pinged by a PR until the tests pass. And an LLM can't claim "Job's Done" before the analysers say the code is OK.
And you don't need to make one yourself, there are tons you can just pick from:
> It's not like testing code is a new thing. Junit is almost 30 years old today.
Unit tests check whether code behaves in specific ways. They certainly are useful to weed out bugs and to ensure that changes don't have unintended side effects.
> And code correctness:
These are tools to check for syntactic correctness. That is, of course, not what I meant.
Algorithmic correctness? Unit tests are great for quickly poking holes in obviously algorithmically incorrect code, but far from good enough to ensure correctness. Passing unit tests is necessary, not sufficient.
Syntactic correctness is more or less a solved problem, as you say. Doesn't matter if the author is a human or an LLM.
It depends on the algorithm of course. If your code is trying to prove P=NP, of course you can't test for it.
But it's disingenuous to claim that even the majority of code written in the world is so difficult algorithmically that it can't be unit-tested to a sufficient degree.
Suppose you're right and the "majority of code" is fully specified by unit testing (I doubt it). The remaining body of code is vast, and the comments in this thread seem to overlook that.
And yet the categorical concepts in Hask are undoubtedly practically useful, more so than an arbitrary sample of concepts, and compose extraordinarily well. Does that have nothing to do with those concepts deriving from (even more general concepts of) category theory?
The claim is furthermore that truly original thinking, the infamous leaps in understanding and creativity, happen by attaining points outside such a convex hull.
It's hard to rigorously verify or disprove this claim. Hopefully this helps build an intuition of why the claim is not as shallow and obviously wrong as it may seem initially.
reply