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

DeepSeek V4 – almost on the frontier

simonwillison.net

Post reply on HN
Print view View original

191–200 of 420 posts

Re: DeepSeek V4 – almost on the frontier

#191

Post by ls612 »

zozbot234 wrote: post #137
Gareth321 wrote: post #135

Earlier quoted context omitted.

Isn't it widely speculated that these are distilled from current frontier models? Distillation is far less compute intensive than primary training. That said, if distillation produces something almost as good for a fraction of the cost, Jensen's point may stand.

You can't really distill a model without access to the internal weights. You could train on chat logs, but that's absolutely not the same thing, it doesn't even come close to comprehensively "extracting" the model's capabilities. And everyone does that in the industry anyway ever since ChatGPT was first released, some versions of Opus even claimed to be DeepSeek if you prompted them in Chinese.

Calling it distillation does however make normies go along with it when they inevitably add all the Chinese labs to the entities list to pad Dario and Sam’s pockets.

Re: DeepSeek V4 – almost on the frontier

#192

Post by Iolaum »

sanex wrote: post #177
oneseven wrote: post #140

Earlier quoted context omitted.

They're probably worried about liability. Let's say that Oracle finds out you reverse engineered their DB using Gemini. You can be sure they will sue Google. Not just for providing the tools, but you could make the argument that it's actually Gemini doing the reverse engineering, and on Google's hardware no less.

We need that lawsuit to happen already so we can establish precedent. The person in the driver's seat of the Tesla should be at fault. The engineer using the llm should be at fault. The person behind the gun not the manufacturer should be at fault.

We shouldn't need a lawsuit. The legislative branch should pass a law clarifying those things, that's their job.

Re: DeepSeek V4 – almost on the frontier

#193

Post by ascorbic »

gordonhart wrote: post #154
sanex wrote: post #138

Earlier quoted context omitted.

We have an enterprise cursor account so I can try all the mainstream models. Using composer 2 on our own code which I obviously have the source code for I couldn't get it to turn on a debug flag to bypass license checks while I was troubleshooting something. Infuriating. It was like that old Patrick from SpongeBob meme. I don't understand why we would turn the models into law enforcement officers. Things that are ill…

> I don't understand why we would turn the models into law enforcement officers It's a simple corporate risk minimization strategy. Just look at how universally despised Grok is on HN. Not because it's a bad model, but because it has less aggressive alignment which means it can be coaxed into saying things that get Xai pilloried here and elsewhere.

No, they've clearly put a lot of work into alignment. It's just that they've been trying to align it with Elon Musk rather than Amanda Askell. Unfortunately the more anti-woke they try to make it, the worse it seems to perform.

Re: DeepSeek V4 – almost on the frontier

#194

Post by 0xbadcafebee »

deaux wrote: post #14

I'm surprised that people here don't care at all about these models openly training on your data, especially if you use them straight from the model developer. Whereas things like "GitHub now automatically opts everyone into using their code for model training" get hundreds of justifiably angry comments, I never see this brought up anymore on posts like these talking about using Chinese models through OpenRouter. Thi…

> I'm surprised that people here don't care at all about these models openly training on your data

You can use zero data retention and zero training providers for most open weights. See OpenRouter and OpenCode Go/Zen for examples.

This is actually one of the big selling points behind open weights - neither China nor the US get your data.

Re: DeepSeek V4 – almost on the frontier

#195

Post by TacticalCoder »

wg0 wrote: post #9

Deepseek v4 Pro feels like Claude Opus 4.6 in it's personality but here's what I did find out about costs: I did cut loose Deepseek v4 on a decent sized Typescript codebase and asked it to only focus on a single endpoint and go in depth on it layer by layer (API, DTOs, service, database models) and form a complete picture of types involved and introduced and ensure no adhoc types are being introduced. It developed a…

> would have burned somewhere between $9 to $13 easily with not much benefit

With not much benefit compared to DeepSeek v4 Pro @ 9 cents (1/100th of the price) or did neither offer any benefit?

Re: DeepSeek V4 – almost on the frontier

#196

Post by noelsusman »

gordonhart wrote: post #154
sanex wrote: post #138

Earlier quoted context omitted.

We have an enterprise cursor account so I can try all the mainstream models. Using composer 2 on our own code which I obviously have the source code for I couldn't get it to turn on a debug flag to bypass license checks while I was troubleshooting something. Infuriating. It was like that old Patrick from SpongeBob meme. I don't understand why we would turn the models into law enforcement officers. Things that are ill…

> I don't understand why we would turn the models into law enforcement officers It's a simple corporate risk minimization strategy. Just look at how universally despised Grok is on HN. Not because it's a bad model, but because it has less aggressive alignment which means it can be coaxed into saying things that get Xai pilloried here and elsewhere.

It's mostly just a bad model. Plenty of people would be willing to overlook the baggage if the model was even marginally better than the competition.

Re: DeepSeek V4 – almost on the frontier

#197

Post by crakhamster01 »

I realize this post is about the pelican test, but in regards to coding, has anyone tried out the advisor strategy with V4?[0]

e.g. Have V4 call out to Opus when it's uncertain, but otherwise handle execution.

The results with Sonnet/Haiku in the blog post seemed promising, so I'm curious how it would go with these latest open models.

[0] https://claude.com/blog/the-advisor-strategy

Re: DeepSeek V4 – almost on the frontier

#198

Post by 0xbadcafebee »

eckelhesten wrote: post #67
dbeley wrote: post #45

Earlier quoted context omitted.

The cool thing about open-weights model is that you are free to use alternative providers that won't phone home to the original model creators. I see 6 alternative providers listed on Openrouter for DeepSeek V4 Pro for example.

At least that’s what they’re telling you. It’s a ”trust me bro” scenario. I’d rather use the phone home version (deepseeks own endpoint). The benefit is that I’m fairly certain that they actually host the model I’m paying for.

If you're not Chinese, and you start a company outside of China, and your whole pitch is "We run open weights and we have nothing to do with China", 1) why would send data to China?? 2) why would you risk your business to do a thing that makes no sense?

Re: DeepSeek V4 – almost on the frontier

#199

Post by driverdan »

prism56 wrote: post #18
deaux wrote: post #14

I'm surprised that people here don't care at all about these models openly training on your data, especially if you use them straight from the model developer. Whereas things like "GitHub now automatically opts everyone into using their code for model training" get hundreds of justifiably angry comments, I never see this brought up anymore on posts like these talking about using Chinese models through OpenRouter. Thi…

If the data is opensource on github, then in my opinion it should be fair game.

The data is not open source. They have open weights but the source data is never open.

Re: DeepSeek V4 – almost on the frontier

#200

Post by gspetr »

deaux wrote: post #72
eckelhesten wrote: post #68

Earlier quoted context omitted.

As opposed to? Do you really think OpenAI, Anthropic or any other entity in the same business respects your data? The Chinese AI companies who release open weights actually deserve whatever input you give them. They are the reason why there is competition and not duopolies in the domain.

I think Google, and likely Anthropic, indeed do honor the settings chosen by the user. For Google in particular it'd be very surprising if they didn't. That's also why both do everything they can to trick users into allowing it. OpenAI, I wouldn't be surprised if you were right.

You mean the same Anthropic, that wouldn't blink an eye at intentionally overcharging users hundreds of dollars just for having a HERMES.md file in a repo, would be above taking your data for... ethical reasons?
Post reply on HN