Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

How I use LLMs to learn complex topics

laurentiugabriel.github.io

Post reply on HN
Print view View original

411–420 of 574 posts

Re: How I use LLMs to learn complex topics

#411

Post by KeplerBoy »

embedding-shape wrote: post #397
KeplerBoy wrote: post #394

Earlier quoted context omitted.

To be fair the entry barrier got a lot lower over the past ~5 years. Now you can write very good CUDA kernels with just a few lines of python DSL code. Zero cpp boilerplate and zero explicit compiler calls. Stuff like Triton, nvidia warp (the language), numba, cupy jax/pallas and so many others really paved the way. You can start out really high-level, run a profiler and then dive deep into the bottlenecks. TL,DR: Ke…

> To be fair the entry barrier got a lot lower over the past ~5 years. Now you can write very good CUDA kernels with just a few lines of python DSL code. Zero cpp boilerplate and zero explicit compiler calls. Well, yeah, but what I've being doing is learning proper CUDA, not "Python-compiled-to-CUDA" (otherwise it'd take like a just a week to understand enough :P ) and that's looking more or less the same today (alth…

I wouldn't call one proper CUDA and the other one some dumbed down version. Nvidia really seems to be pushing for these DSLs to be first class within the ecosystem. In some cases probably even more cutting edge than the nvcc frontend, since it's easier to do some experimenting on a new niche package than on the tool everyone relies on.

I believe more and more production code is running kernels which didn't originate from the traditional cuda cpp route.

Re: How I use LLMs to learn complex topics

#412

Post by latexr »

tempacct2cmmnt wrote: post #74
wxw wrote: post #31

> What you get is a beautiful animation that is 100% accurate and free of hallucinations. I'm not sure I follow how this is actually guaranteed? The fact-checking process mentioned just seems to involve asking AI to review its own work.

Agreed. Given how many significant errors LLMs make in my topic of expertise, despite my taking multiple error checking steps, the idea of catching 100% of hallucinations because you told the LLM to check itself is hilarious. It’s just a wild lack of insight: “I’m using the LLM to teach me something I don’t know about, I definitely have the knowledge base to spot any errors that might remain!”

[dead]

Re: How I use LLMs to learn complex topics

#413

Post by mariusor »

fy20 wrote: post #322
matherial wrote: post #306

This is, without exaggeration, probably the fiftieth blog post or long-form comment about how someone is using an LLM for "complex learning", and I'd just really like to see at least one of these to be accompanied by a statement saying what are the kinds of problems the author can now confidently solve that they couldn't before. In my experience, LLMs are really good for taking up your time and making you feel like y…

I think it depends on what you want to learn and what your definition of "learnt" is. The other day I realised I had no idea how DNA and life works. I guess I studied it at high school (25 years ago), but maybe it didn't go into much detail or it just didn't click. So I asked ChatGPT to explain it to me, I came up with my own mental model from it's explanation, told it that, then it corrected me where I misunderstood…

Can you say in all honesty that a read of the Wikipedia article for DNA would have been less helpful? It's less convenient perhaps, but definitely more authoritative.

Re: How I use LLMs to learn complex topics

#414

Post by endymion-light »

shuwix wrote: post #328
ixsploit wrote: post #311

Earlier quoted context omitted.

I like to use them to generate me tutorial series for technical topics I want to learn or improve. So I start a prompt like this: > I want you to create a tutorial series about X for me. The prime objective is that I improve in topic X so never provide a solution but guide and teach. (for programming never write code). First create a question catalog to assess my current level. Then I would ask it to structure the tu…

I would love to see the "depth" of your knowledge in those topics. You literally proved the OP's point ... thinking you're learning. More like scratching the surface, with lots of invalid data while not being able to recognize what's invalid. It's like with latest vector of attacks being spamming Github with malware injected in proper looking code in hope of AI to index it. Then you paste the code because you don't u…

This feels like an impossible assessment - yes, a model probably can't give you the education that a advanced/expert book on a topic will, but implying that having a verifiable goal is somehow fake learning feels like an intractable problem.

What level of evidence would be sufficient for you to accept that a model may be able to teach a concept?

I'm happy to take on this challenge with a topic of your choosing, but I don't believe there will be an evidence base that satisfies you that the knowledge is earned or deep enough.

Re: How I use LLMs to learn complex topics

#415

Post by curtisblaine »

Those animations are... series of steps, each with a description? How different is this from a plain text list? It's not that you can see forks or that the layout of the assembly line gives you more information. If anything, a plain text list is easier to scroll backwards and forward?

Re: How I use LLMs to learn complex topics

#416

Post by hlovdal »

djaro wrote: post #331
yoyohello13 wrote: post #58

Earlier quoted context omitted.

I’ve found that if I’m not struggling I’m not learning. If stuff is coming fast and easy that’s a sign that what I’m doing is not stretching existing skills enough. It’s true of most things. Running, dieting, weightlifting being uncomfortable is a sign of progress.

This is also why apps like Duolingo don't work for most users. They want to turn language learning into a fun, effortless game but effortless is fundamentally incompatible with learning. Learning is uncomfortable. Reading a difficult (for you) text in a language you don't understand is exhausting and confusing. But that is where improvement happens. Its even worse since Duolingo added a life system (not sure if they…

> If you get everything right, you are not learning.

One of the teachers on the college I went to had a note on his door with "If you understand everything you're doing - you're not learning anything.

Re: How I use LLMs to learn complex topics

#417

Post by gavmor »

aswegs8 wrote: post #382
gavmor wrote: post #68

Earlier quoted context omitted.

Or is it just an unproven attempt? Seems like a pretty straightforward value proposition. Caveat emptor.

If you know anything about investing, you know it is highly likely to be either A a grift or B an innocent, but naive attempt.

I don't know anything about investing, but I know a lot about naivete. :)

Re: How I use LLMs to learn complex topics

#418

Post by akst »

matherial wrote: post #306

This is, without exaggeration, probably the fiftieth blog post or long-form comment about how someone is using an LLM for "complex learning", and I'd just really like to see at least one of these to be accompanied by a statement saying what are the kinds of problems the author can now confidently solve that they couldn't before. In my experience, LLMs are really good for taking up your time and making you feel like y…

RE: On agents taking up your time

One thing I've begun to notice is that LLMs list of a bunch of interesting stuff and raise all these thing you consider but often sometimes you just want a more focused response, so these scatter responses kind of lead you to being overwhelmed and losing focus on what you really wanted to do. At least I've started noticing this.

Like I'd ask about some statical approach taken in a paper and suddenly i'm being bombard with all these potential pivots and things I really need to consider, I kind of just want to consider 1 thing at a time and come to things once I fixed the immediate issue. Sure I have no doubt these other pieces of information are useful but it's just not the most useful information I need right now.

This is less of a problem with coding agents more so putting learning related questions to an LLM, like is the method covered in this paper, yes no? instead I get an exhaustive but overwhelming and indirect response that contains part of the answer. I just wanted to know if it was worth my time going through the paper but now I'm being bombarded told all this tangential information, which is unclear to me if I need to consider right at this moment, it's really distracting.

Maybe this is something others adapted to but i've resorted speaking past it saying, "this is the question please stay on topic" or literally "one thing at a time please" and then it narrows in, but they really stretch your attention thin if you're not more aggressive with keeping them on topic. The smarter models are better, and if you use max compute it does a better job.

I think they can definitely be helpful for learning, but you got play an active role, you can't just consume what it says like content.

Re: How I use LLMs to learn complex topics

#419

Post by TeMPOraL »

fluoridation wrote: post #380
TeMPOraL wrote: post #374

Earlier quoted context omitted.

Self-consistency and consistency with your lived experience are good heuristics. Reality is self-consistent, so anything that doesn't add up indicated an error in the source or your understanding of it (or both). EDIT: I think it's a kind thing you need to tune yourself into. OTOH, I've observed many (most?) people seemingly being completely oblivious to self-consistency issues of their beliefs and mental models, or…

How do you do that with topics not directly experimenceable, like quantum mechanics or, as in this case, silicon manufacture?

The very first thing I mention: self-consistency. It's the only thing you have if you don't have any empirical data. It's the only thing anyone has, really. QM scientists reading QM papers and experiment reports of other people, and talking with each other, are still relying on self-consistency to sniff their own (or other people's) mistakes.

Re: How I use LLMs to learn complex topics

#420

Post by gf000 »

lambda wrote: post #276
voidhorse wrote: post #202

Earlier quoted context omitted.

The "Socratic Method" (aka maieutic) skills annoy me, precisely because when you read them they are the kind of low-effort, low-expertise crap someone who over relies on AI would naively come up with when tasked with the problem of coming up with skills for learning. "Hey the Platonic dialogues are pretty cool and smart, let's do that". The body of literature on learning theory, and beyond that on specific types of l…

Can you provide some references to the research you're referring to? Sounds interesting but you haven't really provided enough information to find it.

Ironically, LLMs are pretty good at finding stuff like that from vague descriptions. With that, I think the parent commenter meant Art Graesser's AutoTutor work based on the terminology used (hint/pump).

TLDR: Actual human tutoring sessions were recorded and analyzed and Socratic questioning was barely used at all. Instead the following pattern was observed:

Pump — "Uh huh?" "What else?" Costs nothing, so try it first.

Hint — points at the region of the answer. "What about the pumpkin's motion sideways?"

Prompt — fishes for one specific word, with the sentence frame supplied. "The pumpkin keeps moving forward at the same ___?"

Assertion — just says it. "It keeps the runner's horizontal velocity."

Post reply on HN