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

Kimi K3: Open Frontier Intelligence

kimi.com

Post reply on HN
Print view View original

871–880 of 1001 posts

Re: Kimi K3: Open Frontier Intelligence

#871

Post by binary132 »

pitched wrote: post #725
katzenversteher wrote: post #718

Earlier quoted context omitted.

Hey Simon, I noticed one thing all LLMs are currently pretty bad at and maybe we could create a benchmark from it. Let an LLM play the role of a dungeon master and tell it to strictly stay in the script/story and only allow realistic player actions. You will notice that they are easily brought off track. E.g. - Tell the LLM that you as a player noticed a strange glow in an NPCs eyes -> the NPC becomes an enemy. - In…

An approach I like to help solving this is antagonistic or review agents. The first agent decides that eye glows turn NPCs into enemies, the second agent is fully dedicated to deciding if that is valid. If the review fails, it leaves notes and the original agent tries again.

If all or most LLMs are susceptible to failing this test, why would an LLM be a good means of evaluating performance on this test without being given a rubric?

Re: Kimi K3: Open Frontier Intelligence

#872

Post by ddxv »

satvikpendem wrote: post #66

Now, will they actually release the weights? Seems like Chinese model providers are slowly closing up, like Alibaba's Qwen 3.6 which did release weights (but not the biggest parameter count ones) and none for 3.7.

The blog post says they will release them July 27

Re: Kimi K3: Open Frontier Intelligence

#873

Post by rhdunn »

katzenversteher wrote: post #718
simonw wrote: post #111

Pelican: https://tools.simonwillison.net/markdown-svg-renderer#url=ht... - rendered via the OpenRouter API: https://openrouter.ai/moonshotai/kimi-k3 95 input, 16,658 output = 25 cents! https://www.llm-prices.com/#it=95&ot=16658&ic=3&oc=15 (13,241 of those were reasoning tokens.) I think that's the most expensive pelican I've rendered through a Chinese model so far.

Hey Simon, I noticed one thing all LLMs are currently pretty bad at and maybe we could create a benchmark from it. Let an LLM play the role of a dungeon master and tell it to strictly stay in the script/story and only allow realistic player actions. You will notice that they are easily brought off track. E.g. - Tell the LLM that you as a player noticed a strange glow in an NPCs eyes -> the NPC becomes an enemy. - In…

Have you tried the https://huggingface.co/LatitudeGames models? They are used by the https://play.aidungeon.com website, but can also be downloaded and used with llama-server in conjunction with something like SillyTavern.

But in general, I've experienced things similar to you. I've also found that LLMs are bad at subtext, e.g. hinting at an NPC being a werewolf or vampire.

Re: Kimi K3: Open Frontier Intelligence

#874

Post by spiderfarmer »

Painsawman123 wrote: post #870

This is the perfect opportunity for anyone who wants to boycott Anthopic and OpenAI!Anyone who cares about the future of Ai has no excuse to continue relying on companies that want to have a monopoly on intelligence.

I wonder if the clique around Peter Thiel still thinks "Competition is for losers!".

Re: Kimi K3: Open Frontier Intelligence

#876

Post by binary132 »

softwaredoug wrote: post #478

So Chinese labs are driving essentially towards commodotized intelligence. Even if its a few months behind the US. Is this a classic 'commoditize my compliment' situation? They want to sell the hardware and infrastructure behind AI and make the software part not the value driver / moat? I can see it. But also even two Chinese labs sinking 100s of millions USD into training isn't exactly commoditization. It's still a…

It’s a nationally-sponsored arms race with Western proprietary AI, so as long as the capital sponsors of Western AI lose their shirts and Chinese LLMs become ubiquitous, I think that qualifies as a win condition. Individual labs are no different than running a local restaurant — yeah, it would be great if you land a massive success story, but just making a bunch of money before moving on to the next thing is cool too, even if it’s just via state-sponsored venture capital.

Re: Kimi K3: Open Frontier Intelligence

#877

Post by dwa3592 »

threepts wrote: post #846
dwa3592 wrote: post #457

This is super exciting. I really need to buy better hardware to try this stuff.

I hope you have a few million dollars, the model requires more than a terabyte of vram unless you lobotomize it a bit.

lobotomizing is what i was thinking. i don't have a million dollars.

Re: Kimi K3: Open Frontier Intelligence

#878

Post by binary132 »

mike_hearn wrote: post #812
Schlagbohrer wrote: post #732

Earlier quoted context omitted.

Google did this by creating Android to undercut Apple. Many US tech firms did this strategy in the 2000s and 2010s of supporting open source alternatives to their opponent's closed source money maker, to undercut the competition. Back then, it let open source have a big boost and we all benefited from that. Hopefully open weights models will do the same so that AI can be more democratized. I wouldn't want to live in…

Android wasn't created to undercut Apple. It was started long before the iPhone project was public. Android was created because Google felt their apps could be huge on mobile (correct) but that mobile operating systems of the time were painful and frustrating to develop for.

oh dear, are we doing repeating the marketing copy of large companies uncritically

Re: Kimi K3: Open Frontier Intelligence

#880

Post by Schlagbohrer »

bottlepalm wrote: post #472
lukebechtel wrote: post #386

> Chip Design > As an early proof of concept, Kimi K3 designed a chip to serve a nano model built on its own architecture. In a single 48-hour autonomous run, K3 built, optimized, and verified the chip using open-source EDA tools on the Nangate 45nm library. Within 4 mm², the chip closes timing at 100 MHz and sustains over 8,700 tokens/s decode throughput in simulation, packing 1.46M standard cells, 0.277 MB of SRAM,…

Really feels like end game type stuff - AI designing its own next versions, designing its own chips, etc.. The advancement is slow, but fast - like a plant growing. We really are the boiling frogs now aren’t we? And the people with eyes wide open are us, and anyone that frequents this site really. Is this Milliways?

Geoffrey Hinton pointed out that exponential growth can be hard to see in the near term (2-3 years) because that part is a bit within predictions, but, it's the exponential part which means 5+ years out is totally unpredictable.
Post reply on HN