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

Kimi K3, Qwen 3.8, and Anthropic's (Potential) Unravelling

emergingtrajectories.com

Post reply on HN
Print view View original

61–70 of 349 posts

Re: Kimi K3, Qwen 3.8, and Anthropic's (Potential) Unravelling

#61

Post by LarsDu88 »

misiti3780 wrote: post #56
LarsDu88 wrote: post #48

The open weight, open architecture releases of the past several days has me more convinced that ultimately, the winner will be whoever burns their models to ASICs fastest. The LLMs themselves are capable of doing some aspects of chip design as evinced by the K3 press release. Furthermore, the frontier models are "good enough" for a wide swathe of tasks and will soon hit that threshold for a good amount of software en…

is anyone doing this ?

The closest example I've seen is ChatJimmy: https://chatjimmy.ai/ a prototype from Taalas running Llama 8B

Scaling this up to 2.8 Trillion (350X increase), will certainly be challenging.

If I was younger and had the right background, I'd love to dive into attempting somethign like this

Re: Kimi K3, Qwen 3.8, and Anthropic's (Potential) Unravelling

#62

Post by carterschonwald »

misiti3780 wrote: post #56
LarsDu88 wrote: post #48

The open weight, open architecture releases of the past several days has me more convinced that ultimately, the winner will be whoever burns their models to ASICs fastest. The LLMs themselves are capable of doing some aspects of chip design as evinced by the K3 press release. Furthermore, the frontier models are "good enough" for a wide swathe of tasks and will soon hit that threshold for a good amount of software en…

is anyone doing this ?

not sure about that, but im actively working on designing ultra sparse models that i want to have perform competitively with stuff 100-10_000 times larger. ehich does yield similar throughput. time will tell id it works out

Re: Kimi K3, Qwen 3.8, and Anthropic's (Potential) Unravelling

#63

Post by eckr »

misiti3780 wrote: post #56
LarsDu88 wrote: post #48

The open weight, open architecture releases of the past several days has me more convinced that ultimately, the winner will be whoever burns their models to ASICs fastest. The LLMs themselves are capable of doing some aspects of chip design as evinced by the K3 press release. Furthermore, the frontier models are "good enough" for a wide swathe of tasks and will soon hit that threshold for a good amount of software en…

is anyone doing this ?

There was a startup that did this for Llama 3, I forgot their name. Etched is also doing some similar things I believe.

Re: Kimi K3, Qwen 3.8, and Anthropic's (Potential) Unravelling

#64

Post by drakythe »

simianwords wrote: post #46
__bjoernd wrote: post #34

Earlier quoted context omitted.

This assumes all the open models are just a result of distilling Anthropic models. Which remains to be proven. And if they are, the point remains that Anthropic has a brittle product advantage that users and investors should be cautious about.

> investors should be cautious about if they are cautious, what would make them invest in newer bigger models without the expected return? generosity?

Sunk cost fallacy.

So much money has been invested in Anthropic and OpenAI at this point that to declare it a loss and walk away could potentially destroy a lot of VC firms, and a non-trivial chunk of the US Economy.

Re: Kimi K3, Qwen 3.8, and Anthropic's (Potential) Unravelling

#65

Post by Incipient »

jhonof wrote: post #28
bko wrote: post #5

I think the risk is overstated. For one, on the margin people are willing to pay a lot for slightly better models. I know personally the value the LLM adds to my workflow is considerably more than the $200/m I pay the frontier labs. I have no interest in optimizing that to get it slightly lower. There are a very vocal minority that optimizes this or companies whose LLM expense is marginal, but I think that's the mino…

> For one, on the margin people are willing to pay a lot for slightly better models. I know personally the value the LLM adds to my workflow is considerably more than the $200/m I pay the frontier labs. I have no interest in optimizing that to get it slightly lower. There are a very vocal minority that optimizes this or companies whose LLM expense is marginal, but I think that's the minority (correct me if I'm wrong,…

$200/mo for 20x Max is great value - but it's not sustainable and won't last forever. Microsoft gave up. Anthropic will too. I use roughly $10k/mo - at that price it's absolutely not worth it.

I'm genuinely not sure where the balancing point even is.

Re: Kimi K3, Qwen 3.8, and Anthropic's (Potential) Unravelling

#67

Post by xur17 »

arrowleaf wrote: post #40
bko wrote: post #5

I think the risk is overstated. For one, on the margin people are willing to pay a lot for slightly better models. I know personally the value the LLM adds to my workflow is considerably more than the $200/m I pay the frontier labs. I have no interest in optimizing that to get it slightly lower. There are a very vocal minority that optimizes this or companies whose LLM expense is marginal, but I think that's the mino…

Here, people tend to forget about enterprise customers. Enterprise is excluded from using these heavily subsidized subscriptions, I know of orgs that are spending around $500k/month for teams of ~100 developers actively using AI.

This is a spot where smaller orgs are getting a definite advantage (both cost and speed). If you're a company with 20 - 50 people, you can probably get away with Claude Teams instead of Claude Enterprise, and pay 10x less in ai costs.

Re: Kimi K3, Qwen 3.8, and Anthropic's (Potential) Unravelling

#68

Post by robgough »

misiti3780 wrote: post #56
LarsDu88 wrote: post #48

The open weight, open architecture releases of the past several days has me more convinced that ultimately, the winner will be whoever burns their models to ASICs fastest. The LLMs themselves are capable of doing some aspects of chip design as evinced by the K3 press release. Furthermore, the frontier models are "good enough" for a wide swathe of tasks and will soon hit that threshold for a good amount of software en…

is anyone doing this ?

obligatory link to https://chatjimmy.ai

Re: Kimi K3, Qwen 3.8, and Anthropic's (Potential) Unravelling

#69

Post by drakythe »

satvikpendem wrote: post #55
drakythe wrote: post #37

Earlier quoted context omitted.

Our current incarnation of capitalism is all about monopolistic behaviors. If these LLM companies do get to the point of being able to replace employees I fully expect them to stop selling shovels and start producing the gold directly, anyone else be damned. And, frankly, this has always been the case. If a product is built on top of another service it has a limited lifespan. Either the product will be purchased or i…

And who will buy the gold? At a certain point everyone will be too poor to buy anything.

No one. I do not believe this to be a good idea. I think its pretty awful and this whole thing, if it goes as far as they keep saying it is going to go, will spin wildly out of control and do an immense amount of damage to people.

But the big money has been skipping Leg Day for ages, and that lack of a foundation will bite us all. Which sucks.

Re: Kimi K3, Qwen 3.8, and Anthropic's (Potential) Unravelling

#70

Post by warm_soup »

Well written article. I liked how the author categorizes companies and their strategies into different buckets (not authors choice of words) and/or combination of harness, data centers, electricity, foundational models. I would have loved to read how companies mentioned are pivoting to build their moat
Post reply on HN