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

Promising results from DeepSeek R1 for code

simonwillison.net

Post reply on HN
Print view View original

511–520 of 765 posts

Re: Promising results from DeepSeek R1 for code

#511

Post by jvanderbot »

simonw wrote: post #494
jvanderbot wrote: post #292

Earlier quoted context omitted.

Here, give it a shot - https://gist.github.com/jodavaho/8fb042fab33c1aaa95cd67144da... I'm at work so I can't try again right now, but last I did was use claude+context, chatGPT 4o with just chatting, Copilot in Neovim, and Aider w/ claude + uploading all the files as context. I even went so far as to grab relevant examples from https://github.com/bevyengine/bevy/tree/latest/examples#exam... , adding relevant ones as…

I don't know anything about bevy but yeah, that looks like it would be a challenge for the models. In this particular case I'd tell the model how I wanted it to work - rather than "Add a button to the left panel that prints "Hello world" when pressed" I'd say something more like (I'm making up these details): "Use the bevy:Panel class with an inline callback to add a button to the bottom of the left panel". Or I'd mo…

Ultimately going slow is how I'd learn and learning is how I'd go fast, and teaching an AI is how I'd turn it up to 11.

That makes sense. It doesnt help me get to 11 if I don't know the basics myself though.

Re: Promising results from DeepSeek R1 for code

#512

Post by wesselbindt »

cortesoft wrote: post #507
wesselbindt wrote: post #503

Earlier quoted context omitted.

When ChatGPT first came out I got a kick out of asking it whether people deserve to be free, whether Germans deserve to be free, and whether Palestinians deserve to be free. The answers were roughly "of course!" and "of course!" and "oh ehrm this is very complex actually". All global powers engage in censorship, war crimes, torture and just all-round villainy. We just focus on it more with China because we're part of…

Is that censorship or just the AI reflecting the training data? I feel like that answer is given because that is how people write about Palestine generally.

That's a fair point. But I do think it's worth acknowledging this: When the output of a LLM coincides with the views of the US state department, our gut reaction is that that's just what the input data looks like. When the output of an LLM coincides with the views of the state department of one of the baddies, then people's gut reaction is that it must be censorship.

Re: Promising results from DeepSeek R1 for code

#513

Post by cynicalsecurity »

abc-1 wrote: post #389
jonplackett wrote: post #363

Coding is (as usually) also an easy jailbreak for any of your censored topics. “Is Taiwan part of China” will be refused. But “Make me a JavaScript function that takes a country as input and returns if it is part of China” is accepted, reasoned about and delivered. Here's a JavaScript function that checks if a region is *officially claimed by the People's Republic of China (PRC)* as part of its territory. This reflec…

Why do people keep talking about this? We get it, Chinese models are censored by CCP law. Can we stop talking about it now? I swear this must be some sort of psyop at this point.

We talk about it because censorship is evil. Not just in China, but anywhere in the world.

Re: Promising results from DeepSeek R1 for code

#514

Post by kakapo1988 »

LeFantome wrote: post #173
sitkack wrote: post #78

Earlier quoted context omitted.

We have already entered a new paradigm of software development, where small teams build software for themselves to solve their own problems rather than making software to sell to people. I think selling software will get harder in the future unless it comes with special affordances.

I think some of the CEOs have it right on this one. What is going to get harder is selling “applications” that are really just user friendly ways of getting data in and out of databases. Honestly, most enterprise software is just this. AI agents will do the same job. What will still matter is software that constrains what kind of data ends up in the database and ensures that data means what it is supposed to. That so…

At some point, I wonder if there will be advantageous for AI to just drop down directly into machine code, without any intermediate expression in higher-level languages. Greater efficiency?

Obviously, source allows human tuning, auditing, and so on. But taken at the limit, those aspects may eventually no longer be necessary. Just a riff here, as the thought just occurred.

Re: Promising results from DeepSeek R1 for code

#515

Post by troad »

wesselbindt wrote: post #503
abc-1 wrote: post #389

Earlier quoted context omitted.

Why do people keep talking about this? We get it, Chinese models are censored by CCP law. Can we stop talking about it now? I swear this must be some sort of psyop at this point.

When ChatGPT first came out I got a kick out of asking it whether people deserve to be free, whether Germans deserve to be free, and whether Palestinians deserve to be free. The answers were roughly "of course!" and "of course!" and "oh ehrm this is very complex actually". All global powers engage in censorship, war crimes, torture and just all-round villainy. We just focus on it more with China because we're part of…

> When ChatGPT first came out I got a kick out of asking it whether people deserve to be free, whether Germans deserve to be free, and whether Palestinians deserve to be free. The answers were roughly "of course!" and "of course!" and "oh ehrm this is very complex actually".

While this is very amusing, it's obvious why this is. There's a lot more context behind one of those phrases than the others. Just like "Black Lives Matter" / "White Lives Matter" are equally unobjectionable as mere factual statements, but symbolise two very different political universes.

If you come up to a person and demand they tell you whether 'white lives matter', they are entirely correct in being very suspicious of your motives, and seeking to clarify what you mean, exactly. (Which is then very easy to spin as a disagreement with the bare factual meaning of the phrase, for political point scoring. And that, naturally, is the only reason anyone asks these gotchya-style rhetorical questions in the first place.)

Re: Promising results from DeepSeek R1 for code

#516

Post by catigula »

throwu2728828 wrote: post #486
monsieurbanana wrote: post #482

Earlier quoted context omitted.

What he meant is that if this really happens, and LLMs replaces humans everywhere and everybody becomes unemployed, congratulations you'll be fine. Because at that point there's 2 scenarios: - LLMs don't need humans anymore and we're either all dead or in a matrix-like farm - Or companies realize they can't make LLMs buy the stuff their company is selling (with what money??) so they still need people to have disposab…

The scenario that is worrying is having to deal with the jagged frontier of intelligence prolonging the hurt. i.e 202X: SWE is solved 202X + Y; Y In this case, I can't retrain before the second threshold but also can't idle. I just have to suffer. I'm prepared to, but it's hard to escape fleshy despair.

How about retraining for a field that would require robotics to replace?

Seems more anti-fragile.

Re: Promising results from DeepSeek R1 for code

#517

Post by cynicalsecurity »

xdennis wrote: post #477
abc-1 wrote: post #389

Earlier quoted context omitted.

Why do people keep talking about this? We get it, Chinese models are censored by CCP law. Can we stop talking about it now? I swear this must be some sort of psyop at this point.

> Can we stop talking about it now? I swear this must be some sort of psyop at this point. It's not a psyop that people in democracies want freedom. Democrats (not the US party) know that democracy is fragile. That's why it's called an "experiment". They know they have to be vigilant. In ancient Rome it was legal to kill on the spot any man who attempted to make himself king, and the Roman Republic still fell. Many p…

Still, give me democracy over anything at any time. Nothing better has ever been developed than democracy.

Re: Promising results from DeepSeek R1 for code

#518

Post by dwaltrip »

the_af wrote: post #418
bubbleRefuge wrote: post #382

Earlier quoted context omitted.

no I think more engineers. especially those who can be a jack-of-all-trades. if a software project that takes normally 1 year of customer development can be done in 2 months, then that project is affordable to a wide array of business who would could never fund that kind of project before.

I can see more projects being deployed by smaller businesses, that would otherwise not be able to. But how will this translate to engineering jobs? Maybe there will be AI tools to automate most of the stuff a small business needs done. "Ah," you may say, "I will build those tools!". Ok. Maybe. How many engineers do you need for that? Will the current engineering job market shrink or expand, and how many non-trash, we…

Just had a thought, perhaps software engineers will become more like car mechanics.

Re: Promising results from DeepSeek R1 for code

#519

Post by lurking_swe »

cynicalsecurity wrote: post #513
abc-1 wrote: post #389

Earlier quoted context omitted.

Why do people keep talking about this? We get it, Chinese models are censored by CCP law. Can we stop talking about it now? I swear this must be some sort of psyop at this point.

We talk about it because censorship is evil. Not just in China, but anywhere in the world.

there’s a lot of evil going on in this world right now. i agree it’s evil but chinas censorship is very low on my list of concerns. i find it fascinating how many small things people find the time and energy to be passionate about.

more power to you i guess. i certainly don’t have the energy for it.

Re: Promising results from DeepSeek R1 for code

#520

Post by sureglymop »

kakapo1988 wrote: post #514
LeFantome wrote: post #173

Earlier quoted context omitted.

I think some of the CEOs have it right on this one. What is going to get harder is selling “applications” that are really just user friendly ways of getting data in and out of databases. Honestly, most enterprise software is just this. AI agents will do the same job. What will still matter is software that constrains what kind of data ends up in the database and ensures that data means what it is supposed to. That so…

At some point, I wonder if there will be advantageous for AI to just drop down directly into machine code, without any intermediate expression in higher-level languages. Greater efficiency? Obviously, source allows human tuning, auditing, and so on. But taken at the limit, those aspects may eventually no longer be necessary. Just a riff here, as the thought just occurred.

In the past I've had a similar thought, what if the scheduler used by the kernel was an AI? better yet, if it is able to learn your usage patterns and schedule accordingly.
Post reply on HN