81–90 of 465 posts
So maybe it makes sense that LLMs do okay at programming even though they lack the ability to reason?
My experience is that they spit out reasonably looking solutions but then they don't even parse/compile.
They are OK to create small spinets of code and completion.
Anything past that they suck.
It's actually hilarious that AI "solved" bullshiting and and artistic fields much better and faster than say reasoning fields like math or programming.
It's the supreme irony. Even 5 years ago the status quo was saying artistic fields were completely safe from the AI apocalypse.
Good code doesn’t just solve a problem, it solves it in a way that’s readable and modular. I think the problem-solving part of coding requires math skills, while the organization part requires writing skills. The organization part affects the problem-solving part, because if you write messy code (that you can’t reread once you forget or extend without rewriting) you’ll quickly get overwhelmed. Writing large math proo…
Most coding doesn't need much of any math past boolean logic and very basic set operations. I'm much more likely to spend my time studying DB and interface schemas to understand how something works than doing a lot of mathy fiddling. Sure, some people write game engines and such, but even much of 3D graphics doesn't need anything more complicated than the first half of a linear algebra course.
So how much of programming is maths? Before we answer that, let's answer: How much of maths is actually maths? Because first we define maths, and then we define programming based on whatever that is, but until we have that first concrete definition this discussion cannot occur.
I will add that "it is taught by the maths department in college" is a flimsy argument, and frankly one the Physics department in particular would mock.
This says more about our programming languages than it does about the brain. I've always wondered why FP isn't more popular. I concluded it's because most folks don't like thinking like abstract math.
I can't speak for other forms of FP, but symbol operators make communicating about haskell very annoying. Outside of that FP seems to be doing fine, IMO.
So maybe it makes sense that LLMs do okay at programming even though they lack the ability to reason?
It's like people enjoy extrapolating their surprise when it comes to LLMs, and I don't think it's very helpful.
Earlier quoted context omitted.
Most coding doesn't need much of any math past boolean logic and very basic set operations. I'm much more likely to spend my time studying DB and interface schemas to understand how something works than doing a lot of mathy fiddling. Sure, some people write game engines and such, but even much of 3D graphics doesn't need anything more complicated than the first half of a linear algebra course.
The difficulty is in how many relationships you need to keep in mind, not in how hard each of them are. Just like in math. BTW relational DBs are math.
It's funny, reading the post you're replying to, I basically read it as
> I don't need math, I need
My teenage daughter used to complain about math, and I spent some time trying to explain to her that we use math every day... EVERY day. Now, when I see her do something that was math (even if it's not obvious it was math), I say "Math... every day". I say that a lot.
Also, yes, my daughter finds me annoying. But also funny; but likely not for the math thing.
There is no such thing as a "language brain" or a "math brain" unless you show experimentally that those bunch of neurons can be grouped into two non-overlapping regions dedicated to "language" and "mathematics". Mathematics itself is a human-made formal language that can be bootstrapped from definitions and axioms of logics and set theory, which have to be given in human language first. Experienced mathematicians re…
Earlier quoted context omitted.
Language has an inherently recursive structure: I saw the man who saw the man who saw the man who saw the man who saw the man who... While our brains have practical limits to how deeply such things can actually be nested, language has a recursive tree-like aspect to it.
Yes, but “language is fundamentally recursive” doesn’t mean the same thing as “recursion is fundamentally linguistic”. Language is just one example of a recursive structure.
Makes sense to me. "Language" is an oft-forgotten part of "programming language." I'm sure math skills probably would better predict, say, ability to do well in advanced theory classes, but programming doesn't require that. It DOES require learning a specialized language.
Which... is what math is, too.