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

GPU World

gpuworld.org

Post reply on HN
Print view View original

91–100 of 303 posts

Re: GPU World

#91

Post by 1718627440 »

bawolff wrote: post #70
rbanffy wrote: post #25

Earlier quoted context omitted.

Most people still go to work every day, inequality still exists, as does famine, disease, wars… Our cars are cleaner, our phones are much better, our TVs much lighter and bigger.

> Most people still go to work every day, Ironically, that is one of the things that have changed. In 1980 working from home was borderline unthinkable. Now, while it is far from universal its also not that unusual.

Society wise yes, but if you had an office job, you essentially take inbound mail and transfrom it into outbound mail. That has been possible to do from home in the 80s. If you work in a large corporation, you also don't talk to people directly, you call them. That's possible from home too.

Until the industrial revolution, every work was home work.

Re: GPU World

#92

Post by dgellow »

maxnevermind wrote: post #77

> Someday, such as in 2040, there may be available, for every human being, the performance equivalent of 'a B300 GPU for contemporary LLMs'. What would this world be like? If we talk about just LLMs, given how things have been going since ChatGPT, my bet it would not change that much. LLMs are not foundational technology such as Internet or Steam engine or Rail roads were. There are very few products that can build u…

Also, way more centralization, with all that logic pushed to a few AI vendors

Re: GPU World

#93

Post by BillStrong »

I mean, why would we wait til then? AMD is buying a company that put LLama 3.1 7b(or something close) on silicon, and can give you 15k t/s, Qwen just released Qwen 3.8 Next that seperates the intelligence from the database, so the intelligence is 6b and the rest can be done on CPU. You can put that 6b on the AMD silicon, and then use another accelerator to speed up the database if you need fast PP, and you are now much faster than the B300 after optimizations such as caching the Prompt once, and keeping it around, and many others.

Suddenly, you can put that on the CPU next to an NPU, or on its own card, on the motherboard or on a GPU.

Re: GPU World

#94

Post by setopt »

maxnevermind wrote: post #77

> Someday, such as in 2040, there may be available, for every human being, the performance equivalent of 'a B300 GPU for contemporary LLMs'. What would this world be like? If we talk about just LLMs, given how things have been going since ChatGPT, my bet it would not change that much. LLMs are not foundational technology such as Internet or Steam engine or Rail roads were. There are very few products that can build u…

I disagree with this take. While LLMs themselves are currently unreliable, the work done in the math community on hooking up creative LLMs to reliable verifiers like Lean show that it’s possible to construct systems where the unreliability is suppressed. For now, that still requires experts to set up and monitor, but I do believe that in a couple of decades we’ll make progress on how to do more mundane tasks in a reliable way without expert supervision, where an LLM still sits as the translation layer between humans and machines. And that universal human-to-machine translator, I would certainly consider a foundational technology.

EDIT: If you asked people on the street in 1990, they’d probably not consider the Internet to be in the same category as the Steam engine either. I mean, you already had phone and fax, so it wasn’t that ground breaking. And I’ve even read articles from the mid-90s declaring the Internet a temporary fad.

Re: GPU World

#95

Post by pjmlp »

They already have, unfortunately most devs still don't target GPU and NPUs available on their mobile devices.

Webviews are naturally not the way to handle them.

Re: GPU World

#96

Post by jstummbillig »

maxnevermind wrote: post #77

> Someday, such as in 2040, there may be available, for every human being, the performance equivalent of 'a B300 GPU for contemporary LLMs'. What would this world be like? If we talk about just LLMs, given how things have been going since ChatGPT, my bet it would not change that much. LLMs are not foundational technology such as Internet or Steam engine or Rail roads were. There are very few products that can build u…

The thing that would change it a lot is that people would start trading away their GPUs, because some people would get vastly more economic use out of them.

Re: GPU World

#97

Post by benfortuna »

jaggederest wrote: post #69
NitpickLawyer wrote: post #15

I just finished reading Service Model by Adrian Tchaikovsky [1], a really timely novel that deals with lots of open ended questions of AI, robots, humanity, control, self determination, and so on. Really recommend it if you're into these kinds of things. > Premise Imagine that AI frontier progress stops as of 1 September 2026: AI becomes faster and cheaper, but it never becomes superhuman or improves considerably acr…

The more I use Fable the more impressed I am with it. I am still learning how to use it, same with gpt-5.6-sol. I don't think the model has changed much, but both models have gotten smarter in leaps and bounds as I've improved the documentation, context, how I provide work to them and how I specify success. (And generally that consists of keeping only the very most relevant details and deleting a lot of the boilerpla…

>I am still learning how to use it

I find it interesting that we are still talking in terms of modifying our behaviour to accommodate the tools. Surely the point of LLMs is not to "learn how to use it" but as an extension of our own unique capabilities - the most personalised tool in all of history.

Re: GPU World

#98

Post by falcor84 »

maxnevermind wrote: post #77

> Someday, such as in 2040, there may be available, for every human being, the performance equivalent of 'a B300 GPU for contemporary LLMs'. What would this world be like? If we talk about just LLMs, given how things have been going since ChatGPT, my bet it would not change that much. LLMs are not foundational technology such as Internet or Steam engine or Rail roads were. There are very few products that can build u…

> it is impossible to describe the universe and compress it into few terabytes

Why? I'm not an expert, but from my understanding, there say species that function well in their ecological niches with significantly smaller cognitive capacities. And we already have autonomous cars and drones navigate and function reasonably well. So I don't see any reason to believe it's impossible. And even if terabytes aren't sufficient, there's no real barrier to increasing capacity.

Re: GPU World

#99

Post by nylonstrung »

raincole wrote: post #44

> What will our world be like when (not if) every human being has access to the equivalent of a Fable or Sol LLM 24/7/365 One single sentence illustrating why tech bro has become a derogatory term. Do they realize that today many humans don't even have access to clean water?

> What will happen in the oft-ignored developing world?

Did you even read the prompt

Re: GPU World

#100

Post by matsemann »

jedberg wrote: post #43
calmoo wrote: post #36

Earlier quoted context omitted.

Pangram is very accurate

No it isn't. I've put human written material in there (my own unpublished material) and it says 90%+ AI. Then I put some AI material in and it said 30% chance of AI. I only tested it with the two pieces, but was not impressed.

Everything has false positives or negatives. 2 examples doesn't tell anything about how good or bad it is.
Post reply on HN