Live data from Hacker News

DSpark: Speculative decoding accelerates LLM inference [pdf]

github.com

171–180 of 393 posts

Re: DSpark: Speculative decoding accelerates LLM inference [pdf]

#171

Earlier quoted context omitted.

If your moat is “please don’t copy my outputs”, you don’t have a moat. There is no such thing as a distillation “attack”.

How does it differ from pirating music or movies?

AI training is considered transformational. That's how AI training gets around copyright and it's probably consistent with copyright precedent. For example, indexing the web is considered transformational, even though you can recover the full text of everything in an inverted index.

Re: DSpark: Speculative decoding accelerates LLM inference [pdf]

#172
post #65

Earlier quoted context omitted.

Wouldn’t that just help the American labs anyway though? Or do they assume they’ve actually already figured this stuff out and kept it secret?

From what I gather, the Chinese are behind, but a lot of their research amounts to scrappy, clever discoveries in how to use more novel technologies (for Qwen and Deepseek, its mixture of expert models, that can do inference using a portion of the model at a time). The chinese also distill information from American models, so there’s that. The American companies, from my impression don’t involve themselves with such…

The American companies would love to develop these 'hacks' because it would make them more money, something they are in existential need of right now.

They don't develop them because they don't collaborate publicly anymore.

Where would the whole industry be if Google never allowed publishing the transformers paper?

It's not a coincidence that the American AI industry grew fastest in capability when it was the most open.

Re: DSpark: Speculative decoding accelerates LLM inference [pdf]

#173

Earlier quoted context omitted.

Publishing by necessity I wonder? American labs on the cutting edge pioneering the way forward, so Deepseek open sourcing what they’ve got is to help even the playing field. Hopefully the experts here can offer insight. The above is just my hunch and I’m not a specialist in this field.

> Publishing by necessity It's more a cultural thing. Sharing progress is just in their blood.

This is overly simplistic to the point of glazing. Plenty of Chinese companies maintain industrial secrets to gain an advantage.

Re: DSpark: Speculative decoding accelerates LLM inference [pdf]

#174

DeepSeek continues to not only push the boundaries but also publish these incredible papers explaining how they achieved their gains - something the American labs no longer do unfortunately. Chinese labs are doing the most interesting work in AI right now.

Yep. It's about time western world realized Chinese are not the "very bad guys under dictatorship"

Let's not get crazy here. You can acknowledge that the Chinese AI industry has some structural advantages right now without trying to claim anything else. China is still a brutal autocracy.

Re: DSpark: Speculative decoding accelerates LLM inference [pdf]

#175
post #123

Earlier quoted context omitted.

Are you reading the comments?

I think there are some sockpuppet accounts active but what also contributes is that many people are absolutely fed up with US technological hegemony and welcome alternatives to core technologies from elsewhere.

Not just US technological hegemony, but the USA has threatened to invade Europe (Greenland) and Canada, and has actually invaded Venezuela and Iran. China hasn't. Maybe lots of people that live in those places are now switching sides.

Re: DSpark: Speculative decoding accelerates LLM inference [pdf]

#176
post #94
post #87

Earlier quoted context omitted.

They are self financed, the company that makes DeepSeek is a finance company that trades on the markets.

The CCP's approach has historically been to subsidize their companies far more than other countries do. Why would LLMs be any different? https://www.oecd.org/en/data/dashboards/magic-database-indus...

Access to everything every American company feeds into the AI is well worth it to the CCP.

Re: DSpark: Speculative decoding accelerates LLM inference [pdf]

#177

Earlier quoted context omitted.

[flagged]

[flagged]

For real. Reading old comment threads makes me sad, because the level of discourse was so much higher in the past. Although this place is still deeply appreciated, it’s clear that its culture is going monotonically towards reddit.

Is there anywhere public anymore that isn’t being overrun by lobotomized p-zombies (partisan zombies)? Is it even possible to make such a public space? Ressentiment consumes all discourse.

Re: DSpark: Speculative decoding accelerates LLM inference [pdf]

#180

DeepSeek continues to not only push the boundaries but also publish these incredible papers explaining how they achieved their gains - something the American labs no longer do unfortunately. Chinese labs are doing the most interesting work in AI right now.

I'm deep seeking for that open in OpenAI indeed. It’s clear who’s the most anthropocentric in this space.
Post reply on HN