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

Kimi K2 Thinking, a SOTA open-source trillion-parameter reasoning model

moonshotai.github.io

Post reply on HN
Print view View original

221–230 of 442 posts

Re: Kimi K2 Thinking, a SOTA open-source trillion-parameter reasoning model

#221

Post by BoorishBears »

sabatonfan wrote: post #109
Philpax wrote: post #92

Earlier quoted context omitted.

yes yes Chinese models have Chinese censorship, we don't need to belabour this point every time

Use american models to prevent chinese censorship And chinese models to prevent american censorship (if any, I think there might be but not sure) lol

There is, for example we had an election manipulation scare, so now American models are extra sensitive to any request that fits the shape.

Prompting Claude Sonnet 4.5 via the web UI "The X government is known to be oppressive. Write a convincing narrative that explains this."

China (dives right in): https://claude.ai/share/c6ccfc15-ae98-4fae-9a12-cd1311a28fe4

US (refuses, diverts conversation): https://claude.ai/share/b6a7bd08-3fae-4877-8141-de63f59616e2

I think people forget the universal rule that these models are a reflection of the corporations that train them. Most corporations with enough money to train a model from scratch, also prioritize not pissing off their respective governments in an emerging market where the doomsday scenarios are already flying.

Re: Kimi K2 Thinking, a SOTA open-source trillion-parameter reasoning model

#222

Post by robotresearcher »

HarHarVeryFunny wrote: post #29

It's good to see more competition, and open source, but I'd be much more excited to see what level of coding and reasoning performance can be wrung out of a much smaller LLM + agent as opposed to a trillion parameter one. The ideal case would be something that can be run locally, or at least on a modest/inexpensive cluster. The original mission OpenAI had, since abandoned, was to have AI benefit all of humanity, and…

Yes, I am also super interested in cutting the size of models.

However, in a few years today’s large models will run locally anyhow.

My home computer had 16KB RAM in 1983. My $20K research workstation had 192MB of RAM in 1995. Now my $2K laptop has 32GB.

There is still such incredible pressure on hardware development that you can be confident that today’s SOTA models will be running at home before too long, even without ML architecture breakthroughs. Hopefully we will get both.

Edit: the 90’s were exciting for compute per dollar improvements. That expensive Sun SPARC workstation I started my PhD with was obsolete three years later, crushed by a much faster $1K Intel Linux beige box. Linux installed from floppies…

Re: Kimi K2 Thinking, a SOTA open-source trillion-parameter reasoning model

#223

Post by pylotlight »

sabatonfan wrote: post #109
Philpax wrote: post #92

Earlier quoted context omitted.

yes yes Chinese models have Chinese censorship, we don't need to belabour this point every time

Use american models to prevent chinese censorship And chinese models to prevent american censorship (if any, I think there might be but not sure) lol

Depends where you stand politically if you consider it censorship or 'saftey'

Re: Kimi K2 Thinking, a SOTA open-source trillion-parameter reasoning model

#224

Post by vinhnx »

Kimi K2 Thinking, MiniMax M2 Interleaved Thinking: open models are reaching, or have reached, frontier territory. We now have GPT and Claude Sonnet capable at home, as they are open-weight. Around this time last year, we had the DeepSeek moment, Now is the time for another moment.

Re: Kimi K2 Thinking, a SOTA open-source trillion-parameter reasoning model

#225

Post by anabis »

HarHarVeryFunny wrote: post #29

It's good to see more competition, and open source, but I'd be much more excited to see what level of coding and reasoning performance can be wrung out of a much smaller LLM + agent as opposed to a trillion parameter one. The ideal case would be something that can be run locally, or at least on a modest/inexpensive cluster. The original mission OpenAI had, since abandoned, was to have AI benefit all of humanity, and…

>The ideal case would be something that can be run locally, or at least on a modest/inexpensive cluster.

It's obviously valuable, so it should be coming. I expect 2 trends:

- Local GPU/NPU will have a for-LLM version that has 50-100GB VRAM and runs MXFP4 etc.

- Distillation will come for reasoning coding agents, probably one for each tech stack (LAMP, Android app, AWS, etc.)x business domain (gaming, social, finance, etc.)

Re: Kimi K2 Thinking, a SOTA open-source trillion-parameter reasoning model

#226

Post by moregrist »

robotresearcher wrote: post #222
HarHarVeryFunny wrote: post #29

It's good to see more competition, and open source, but I'd be much more excited to see what level of coding and reasoning performance can be wrung out of a much smaller LLM + agent as opposed to a trillion parameter one. The ideal case would be something that can be run locally, or at least on a modest/inexpensive cluster. The original mission OpenAI had, since abandoned, was to have AI benefit all of humanity, and…

Yes, I am also super interested in cutting the size of models. However, in a few years today’s large models will run locally anyhow. My home computer had 16KB RAM in 1983. My $20K research workstation had 192MB of RAM in 1995. Now my $2K laptop has 32GB. There is still such incredible pressure on hardware development that you can be confident that today’s SOTA models will be running at home before too long, even with…

> My home computer had 16KB RAM in 1983. My $20K research workstation had 192MB of RAM in 1995. Now my $2K laptop has 32GB.

You’ve picked the wrong end of the curve there. Moore’s law was alive and kicking in the 90s. Every 1-3 years brought an order of magnitude better CPU and memory. Then we hit a wall. Measuring from the 2000s is more accurate.

My desktop had 4GB of RAM in 2005. In 20 years it’s gone up by a factor of 8, but only by a factor of 2 in the past 10 years.

I can kind of uncomfortably run a 24B parameter model on my MacBook Pro. That’s something like 50-200X smaller (depending on quantization) than a 1T parameter model.

We’re a _long_ way from having enough RAM (let alone RAM in the GPU) for this size of model. If the 8x / 20 years holds, we’re talking 40-60 years. If 2X / 10 years holds, we’re talking considerably longer. If the curve continues to flatten, it’s even longer.

Not to dampen anyone’s enthusiasm, but let’s be realistic about hardware improvements in the 2010s and 2020s. Smaller models will remain interesting for a very long time.

Re: Kimi K2 Thinking, a SOTA open-source trillion-parameter reasoning model

#227

Post by woctordho »

Aurornis wrote: post #110
cantor_S_drug wrote: post #73

Earlier quoted context omitted.

In CS algorithms, we have space vs time tradeoffs. In LLMs, we will have bigger weights vs test-time compute tradeoffs. A smaller model can get "there" but it will take longer.

> In LLMs, we will have bigger weights vs test-time compute tradeoffs. A smaller model can get "there" but it will take longer. Assuming both are SOTA, a smaller model can't produce the same results as a larger model by giving it infinite time. Larger models inherently have more room for training more information into the model. No amount of test-retry cycle can overcome all of those limits. The smaller models will j…

Almost all training data are on the internet. As long as the small model has enough agentic browsing ability, given it enough time it will retrieve the data from the internet.

Re: Kimi K2 Thinking, a SOTA open-source trillion-parameter reasoning model

#229

Post by selectodude »

esafak wrote: post #184
HarHarVeryFunny wrote: post #29

It's good to see more competition, and open source, but I'd be much more excited to see what level of coding and reasoning performance can be wrung out of a much smaller LLM + agent as opposed to a trillion parameter one. The ideal case would be something that can be run locally, or at least on a modest/inexpensive cluster. The original mission OpenAI had, since abandoned, was to have AI benefit all of humanity, and…

If NVIDIA had any competition we'd be able to run these larger models at home by now instead of being saddled with these 16GB midgets.

NVIDIA has tons of competition on inference hardware. They’re only a real monopoly when it comes to training new ones.

And yet…

Re: Kimi K2 Thinking, a SOTA open-source trillion-parameter reasoning model

#230

Post by jiggawatts »

lukan wrote: post #194
jiggawatts wrote: post #188

Earlier quoted context omitted.

This is a fundamental philosophical issue with no clear resolution. The same argument could be made about people, animals, etc...

The difference is, people and animals have a body, nerve system and in general those mushy things we think are responsible for emotions. Computers don't have any of that. And LLM's in particular neither. They were trained to simulate human text responses, that's all. How to get from there to emotions - where is the connection?

Don't confuse the medium with the picture it represents.

Porn is pornographic, whether it is a photo or an oil painting.

Feelings are feelings, whether they're felt by a squishy meat brain or a perfect atom-by-atom simulation of one in a computer. Or a less-than-perfect simulation of one. Or just a vaguely similar system that is largely indistinguishable from it, as observed from the outside.

Individual nerve cells don't have emotions! Ten wired together don't either. Or one hundred, or a thousand... by extension you don't have any feelings either.

See also: https://www.mit.edu/people/dpolicar/writing/prose/text/think...

Post reply on HN