Earlier quoted context omitted.
> It makes no sense at all, it's cuckooland, are you all on crazy pills? Frequent LLM usage impairs thinking. The LLM has no connection to reality, and it takes over people's minds.
>Frequent LLM usage impairs thinking Is there hard evidence on this?
Andrej Karpathy: Software in the era of AI [video]
731–740 of 827 posts
Re: Andrej Karpathy: Software in the era of AI [video]
#732Earlier quoted context omitted.
> It makes no sense at all, it's cuckooland, are you all on crazy pills? Frequent LLM usage impairs thinking. The LLM has no connection to reality, and it takes over people's minds.
>Frequent LLM usage impairs thinking Is there hard evidence on this?
If you are the type who prefers studies:
https://time.com/7295195/ai-chatgpt-google-learning-school/
Otherwise, read pro-LLM blogs which are mostly rambling nonsense that overpromises while almost no actual LLM written software exists.
You can also see how the few open source developers who jump on the LLM bandwagon now have worse blogging and programming output than they had pre-LLM.
Re: Andrej Karpathy: Software in the era of AI [video]
#733Earlier quoted context omitted.
An output of a fab, just like an output of an LLM, is non-deterministic, but is good enough, or is being optimized to be good enough. Non-determinism is not the problem, it's the quality of the software that matters. You can repeatedly ask me to solve a particular leetcode puzzle, and every time I might output a slightly different version. That's fine as long as the code solves the problem. The software running on th…
Hardware always involves some level of non-determinism, because the physical world is messier than the virtual software world. Every hardware engineer accepts that and learns how to design solutions despite those constraints. But you're right, non-determinism is not the current problem in some fabs, because the whole process has been modeled with it in mind, and it's the yield ratio that needs to be deterministic eno…
The output of the verilog optimizer is different every time. The output of a fab is different in every batch. Each chip in a batch is different from others in that batch. Quality control drops the fraction of truly poor chips, and hardware design features might downgrade some of the partially failed chips to be classified as lesser versions of the same initial design. The final chips work as intended, mostly, but perhaps the error tolerance to overclocking or the mean time between failures is slightly different between chips. We can all work with them just fine almost all the time. The same principles apply to complex LLM-orchestrated code projects. I dont mind if my compiler gives different code each time because it uses a stochastic optimizer, but I want my code to do what I want and to not fail more than a certain tolerance I have for this code, which depends on the application. By giving more insight into the layers of testing to more people, and by encouraging the new documentation practices that Andrej mentioned, LLM coding will change the practice of software engineering rather dramatically. Code 2.0 was flexible and could yield results that were better than human coded efforts for complex problems, but the architecture, code, data, were selected by humans. In code 3.0 humans have access to (non-deterministic) building blocks that are written in natural language, to bug fixes and feature addition that happen in a conversation style. Similar engineering principles as with code 1.0 still apply (even more so than with code2.0, unless the product is a neural net), but the emphasis on verification increased dramatically as a fraction of the total effort, even though the total effort has gone down a lot. I can’t wait to see increased help in code verification efforts from this batch of people in the AI startup school as a result of Andrej’s presentation.
Re: Andrej Karpathy: Software in the era of AI [video]
#734Earlier quoted context omitted.
> It makes no sense at all, it's cuckooland, are you all on crazy pills? Frequent LLM usage impairs thinking. The LLM has no connection to reality, and it takes over people's minds.
>Frequent LLM usage impairs thinking Is there hard evidence on this?
Some preliminary evidence:
Re: Andrej Karpathy: Software in the era of AI [video]
#735I'm a little surprised at how negative he is towards textual interfaces and text for representing information.
I didn't get the impression that he's against text per se, just that LLMs should use a format that's most concise for humans in the given scenario. Example from the video: showing the (textual) diff between old and new versions of text/code, rather than just the new version. Or converting a text-only restaurant menu to photos+text.
Re: Andrej Karpathy: Software in the era of AI [video]
#736I'm not sure about the 1.0/2.0/3.0 classification, but it did lead me to think about LLMs as a programming paradigm: we've had imperative & declarative, procedural & functional languages, maybe we'll come to view deterministic vs. probabilistic (LLMs) similarly. def __main__: You are a calculator. Given an input expression, you compute the result and print it to stdout, exiting 0. Should you be unable to do this, you…
[flagged]
Do you think you could condense your point of view without hyperbole and rudeness so the rest of us can understand it?
Re: Andrej Karpathy: Software in the era of AI [video]
#737Earlier quoted context omitted.
[flagged]
> It makes no sense at all, it's cuckooland, are you all on crazy pills? Frequent LLM usage impairs thinking. The LLM has no connection to reality, and it takes over people's minds.
Sounds like you’re taking crazy pills.
Far to early from any of the studies done so far to come to your conclusion.
Re: Andrej Karpathy: Software in the era of AI [video]
#738Earlier quoted context omitted.
> It makes no sense at all, it's cuckooland, are you all on crazy pills? Frequent LLM usage impairs thinking. The LLM has no connection to reality, and it takes over people's minds.
>Frequent LLM usage impairs thinking Is there hard evidence on this?
Short answer no.
Longer answer there was that study posted this week that compared it to using search and then what was it…raw thinking or something similar. I could totally understand in certain cases you are not activating parts of your brain as much, I don’t know any of it proves much in aggregate.
Re: Andrej Karpathy: Software in the era of AI [video]
#739Earlier quoted context omitted.
Why do people keep having this reaction to something we're already used to? When you're developing against an API, you're already doing the same thing, planning for what happens when the request hangs, or fails completely, or gives a different response, and so on. Same for basically any IO. It's almost not even new, just that it generates text instead of JSON, or whatever. But we've already been doing "indeterministi…
You’re right about the trees but wrong (hear me out) about the forest. Yes, programming isn’t always deterministic, not just due to the leftpad API endpoint being down, but by design - you can’t deterministically tell which button the user is going to click. So far so good. But, you program for the things that you expect to happen, and handle the rest as errors. If you look at the branching topology of well-written c…
Programmers aren't deterministic either. If I ask ten programmers to come up with a solution to the same problem, I'm not likely to get ten identical copies. Different programmers, even competent experienced programmers, might have different priorities that aren't in the requirements. For example, trading off program maintainability or portability over performance.
The same could apply to LLMs, or even different runs from the same LLMs.
Re: Andrej Karpathy: Software in the era of AI [video]
#740I'm not sure about the 1.0/2.0/3.0 classification, but it did lead me to think about LLMs as a programming paradigm: we've had imperative & declarative, procedural & functional languages, maybe we'll come to view deterministic vs. probabilistic (LLMs) similarly. def __main__: You are a calculator. Given an input expression, you compute the result and print it to stdout, exiting 0. Should you be unable to do this, you…
LLMs are not inherently indeterministic. Batching, temperature, and other things make them appear so when run by big providers but a locally-run LLM model at zero temperature will always produce the same output given the same input.