Promising results from DeepSeek R1 for code • Hacker News Board

Re: Promising results from DeepSeek R1 for code

#111

Post by Espressosaurus » Tue, Jan 28, 2025, 4:52 PM UTC

SecretDreams wrote: post #92

simonw wrote: post #52

Earlier quoted context omitted.

"Jobs are going to be lost unless there's somehow a demand for more applications." That's why I'm not worried. There is already SO MUCH more demand for code than we're able to keep up with. Show me a company that doesn't have a backlog a mile long where most of the internal conversations are about how to prioritize what to build next. I think LLM assistance makes programmers significantly more productive, which makes…

The big fear shouldn't be on loss of jobs, it should be the inevitable attack on wages. Wage will track inversely to proximity as a commodity status. Even the discussion around AI partially replacing coders is a direction towards commoditization.

It's the same thing. If there are more workers than jobs, wages go down. If there are more jobs than workers, wages go up.

We saw it crystal clear between the boom years, the trough, and the current recovery.

Re: Promising results from DeepSeek R1 for code

#112

Post by simonw » Tue, Jan 28, 2025, 4:52 PM UTC

jvanderbot wrote: post #103

jeswin wrote: post #31

> 99% of the code in this PR [for llama.cpp] is written by DeekSeek-R1 I hope we can put to rest the argument that LLMs are only marginally useful in coding - which are often among the top comments on many threads. I suppose these arguments arise from (a) having used only GH copilot which is the worst tool, or (b) not having spent enough time with the tool/llm, or (c) apprehension. I've given up responding to these.…

This is great. Really! Buuut... How do you get these tools to not fall over completely when relying on an existing non-public codebase that isn't visible in just the current file? Or, how do you get them to use a recent API that doesn't dominate their training data? Combining the both, I just cannot for the life of me get them to be useful beyond the most basic boilerplate. Arguably, SIMD intrinsics are a one-to-one…

"Or, how do you get them to use a recent API that doesn't dominate their training data?"

Paste in the documentation or some examples. I do this all the time - "teaching" an LLM about an API it doesn't know yet is trivially easy if you take advantage of the longer context inputs to models these days.

Re: Promising results from DeepSeek R1 for code

#113

Post by anotherpaulg » Tue, Jan 28, 2025, 4:52 PM UTC

joshstrange wrote: post #106

anotherpaulg wrote: post #61

> 99% of the code in this PR [for llama.cpp] is written by DeekSeek-R1 It's definitely possible for AI to do a large fraction of your coding, and for it to contribute significantly to "improving itself". As an example, aider currently writes about 70% of the new code in each of its releases. I automatically track and share this stat as graph [0] with aider's release notes. Before Sonnet, most releases were less than…

First off I want to thank you for Aider. I’ve had so much fun playing with it and using it for real work. It’s an amazing tool. How do you determine how much was written by you vs the LLM? I assume it consists of parsing the git log and getting LoC from that or similar? If the scripts are public could you point me at them? I’d love to run it on a recent project I did using aider.

Glad to hear you’re finding aider useful!

There’s a faq entry about how these stats are computed [0]. Basically using git blame, since aider is tightly integrated with git.

The faq links to the script that computes the stats. It’s not designed to be used on any repo, but you (or aider) could adapt it.

You’re not the first to ask for these stats about your own repo, so I may generalize it at some point.

[0] https://aider.chat/docs/faq.html#how-are-the-aider-wrote-xx-...

Re: Promising results from DeepSeek R1 for code

#114

Post by nkozyra » Tue, Jan 28, 2025, 4:55 PM UTC

jeswin wrote: post #31

> 99% of the code in this PR [for llama.cpp] is written by DeekSeek-R1 I hope we can put to rest the argument that LLMs are only marginally useful in coding - which are often among the top comments on many threads. I suppose these arguments arise from (a) having used only GH copilot which is the worst tool, or (b) not having spent enough time with the tool/llm, or (c) apprehension. I've given up responding to these.…

> it isn't entirely bluster

"Development" is effectively translating abstractions of an intended operation to machine language.

What I find kind of funny about the current state is we're using large language models to, like, spit out React or Python code. This use case is obviously an optimization to WASM, so a little closer to the metal, but at what point to programs (effectively suites of operations) just cut out the middleman entirely?

Re: Promising results from DeepSeek R1 for code

#115

Post by ngxson » Tue, Jan 28, 2025, 4:56 PM UTC

Hi I'm Xuan-Son,

Small correct, I'm not just asking it to convert ARM NEON to SIMD, but for the function handling q6_K_q8_K, I asked it to reinvent a new approach (without giving it any prior examples). The reason I did that was because it failed writing this function 4 times so far.

And a bit of context here, I was doing this during my Sunday and the time budget is 2 days to finish.

I wanted to optimize wllama (wasm wrapper for llama.cpp that I maintain) to run deepseek distill 1.5B faster. Wllama is totally a weekend project and I can never spend more than 2 consecutive days on it.

Between 2 choices: (1) to take time to do it myself then maybe give up, or (2) try prompting LLM to do that and maybe give up (at worst, it just give me hallucinated answer), I choose the second option since I was quite sleepy.

So yeah, turns out it was a great success in the given context. Just does it job, saves my weekend.

Some of you may ask, why not trying ChatGPT or Claude in the first place? Well, short answer is: my input is too long, these platforms straight up refuse to give me the answer :)

Re: Promising results from DeepSeek R1 for code

#116

Post by nkozyra » Tue, Jan 28, 2025, 4:57 PM UTC

matsemann wrote: post #97

jeswin wrote: post #31

> 99% of the code in this PR [for llama.cpp] is written by DeekSeek-R1 I hope we can put to rest the argument that LLMs are only marginally useful in coding - which are often among the top comments on many threads. I suppose these arguments arise from (a) having used only GH copilot which is the worst tool, or (b) not having spent enough time with the tool/llm, or (c) apprehension. I've given up responding to these.…

I read that Meta is tasking all engineers with figuring out how they got owned by deepseek. Couldn't they just have asked an llm instead? After their claim of replacing all of us... I'm not too worried. If anything we're the last generation that knows how to debug and work through issues.

> If anything we're the last generation that knows how to debug and work through issues.

I suspect that comment might soon feel like saying "not too worried about assembly line robots, we're the only ones who know how to screw on the lug nuts when they pop off"

Re: Promising results from DeepSeek R1 for code

#117

Post by tomrod » Tue, Jan 28, 2025, 4:57 PM UTC

onetimeusename wrote: post #107

jeswin wrote: post #86

Earlier quoted context omitted.

> That's why I'm not worried. There is already SO MUCH more demand for code than we're able to keep up with. Show me a company that doesn't have a backlog a mile long where most of the internal conversations are about how to prioritize what to build next. I worry about junior developers. It will be a while before vocational programming courses retool to teach this new way of writing code, and these are going to be te…

Just as a side note, at my university about half the CS people are in the AI track. I would guess that number will keep increasing. There is also a separate major that kind of focuses on AI/psychology that is pretty popular but I am not sure how many people are in it. A good number of the students have some kind of "AI startup". Also, although it violates the honor code, I would be willing to bet many students use AI…

Yeah, the younguns smell opportunity and run towards it. They'll be fine. It's younguns) the less experienced folks in the current corporate world that will have the most to lose.

Re: Promising results from DeepSeek R1 for code

#118

Post by jhsvsmyself » Tue, Jan 28, 2025, 4:58 PM UTC

jeswin wrote: post #86

simonw wrote: post #52

Earlier quoted context omitted.

"Jobs are going to be lost unless there's somehow a demand for more applications." That's why I'm not worried. There is already SO MUCH more demand for code than we're able to keep up with. Show me a company that doesn't have a backlog a mile long where most of the internal conversations are about how to prioritize what to build next. I think LLM assistance makes programmers significantly more productive, which makes…

> That's why I'm not worried. There is already SO MUCH more demand for code than we're able to keep up with. Show me a company that doesn't have a backlog a mile long where most of the internal conversations are about how to prioritize what to build next. I worry about junior developers. It will be a while before vocational programming courses retool to teach this new way of writing code, and these are going to be te…

There are already courses that are centered around coding with AI.

Re: Promising results from DeepSeek R1 for code

#119

Post by menaerus » Tue, Jan 28, 2025, 4:58 PM UTC

simonw wrote: post #99

UncleOxidant wrote: post #95

Earlier quoted context omitted.

> That's why I'm not worried. There is already SO MUCH more demand for code than we're able to keep up with. Show me a company that doesn't have a backlog a mile long where most of the internal conversations are about how to prioritize what to build next. And yet many companies aren't hiring developers right now - folks in the C suite are thinking AI is going to be eliminating their need to hire engineers. Also "dema…

I continue to suspect that the hiring problems are mainly due to massive over-hiring during Covid, followed by layoffs that flooded the market with skilled developers looking for work. I'd love to see numbers around the "execs don't think they need engineers because of AI" factor. I've heard a few anecdotal examples of that but it's hard to tell if it's a real trend or just something that catches headlines.

Massive overhiring or not, it's the fact that many (skilled) engineers can't find a job. Many companies were shut off during the past few years and market became oversaturated over the night. Whether AI will help to correct the market creating more demand we will see but I wouldn't hold my breath. Many domain specific skills became a commodity.

Re: Promising results from DeepSeek R1 for code

#120

Post by AnotherGoodName » Tue, Jan 28, 2025, 4:58 PM UTC

onetimeusename wrote: post #107

jeswin wrote: post #86

Earlier quoted context omitted.

> That's why I'm not worried. There is already SO MUCH more demand for code than we're able to keep up with. Show me a company that doesn't have a backlog a mile long where most of the internal conversations are about how to prioritize what to build next. I worry about junior developers. It will be a while before vocational programming courses retool to teach this new way of writing code, and these are going to be te…

Just as a side note, at my university about half the CS people are in the AI track. I would guess that number will keep increasing. There is also a separate major that kind of focuses on AI/psychology that is pretty popular but I am not sure how many people are in it. A good number of the students have some kind of "AI startup". Also, although it violates the honor code, I would be willing to bet many students use AI…

The ask from every new grad to be assigned to ai development is unreasonable right now and they are probably hurting their careers by all going the same direction honestly. It’s a small fraction of our development efforts and we usually hire very senior for that sort of role. We still need people that can program for the day to day business needs and it’s a perfect starting role for a new grad yet almost all of them are asking for assignment to ai development.

I appreciate anyone that can utilise ai well but there’s just not enough core ai model development jobs for every new grad.