Hacker News: segmondy

New comment by segmondy in "Kimi-K3 Technical Report [pdf]"

segmondy — Mon, 27 Jul 2026 16:37:18 +0000

if it's to be believed, it's so easy. you distill it, and you have a copy in 2 weeks, isn't that what China is doing? so 2 weeks from now, we should have one.

New comment by segmondy in "Open-weight AI is having its Kubernetes moment"

segmondy — Sat, 25 Jul 2026 19:27:07 +0000

I don't know what you mean by "economical", but it has been "economical" to run this stuff yourself for the last 3 years.

1. You must be willing to be resourceful. 2. Be willing to learn, do the hard things. 3. Accept the tradeoffs.

New comment by segmondy in "Open-weight AI is having its Kubernetes moment"

segmondy — Sat, 25 Jul 2026 19:25:13 +0000

What is strange about GPT-4 being expensive in 2023? Supply and demand. Which other model choices did we have? Prices are related to supply and demand. We see it play out with the introduction of capable open weight models or even other closed cloud models.

New comment by segmondy in "Nvidia, Microsoft, Meta warn against overregulating open-weight models"

segmondy — Fri, 24 Jul 2026 17:18:40 +0000

At the end of the day, these companies that signed this letter have more cash. What they should do is bring some money together, more money than the other side, donate it and buy their way. That's the 2026 way, no one has time for all this letter and protest. How much are they willing to put on the line?

New comment by segmondy in "Nvidia, Microsoft, Meta warn against overregulating open-weight models"

segmondy — Fri, 24 Jul 2026 17:17:01 +0000

Google has produced Gemini Pro, Meta produced llama3-70B/llama3-405B and now has Muse, Musk has Grok. These are very capable models and to claim that they are not frontier is to stretch the meaning unless you want to say "#1 model" Gemini Pro at one point even if for a week, was #1 model.

New comment by segmondy in "Nvidia, Microsoft, Meta warn against overregulating open-weight models"

segmondy — Fri, 24 Jul 2026 17:14:21 +0000

ha! who is groveling to the white house to stop competition? openai, anthropic and google, they are the losers. if they want to compete, compete.

New comment by segmondy in "Kimi K3 Is Competitive with Fable; Kimi K3 and Fable Is SoTA"

segmondy — Wed, 22 Jul 2026 02:20:47 +0000

This is pure speculation. China labs are geeks and nerds like some of us. They were amazed at LLM and like everyone want to build their own. They were happy to get meaningful next token predictions. I think most of you forgot how bad these things were 3 years ago compared to today. There was nothing to undercut, it was just geeks putting out their toys and saying, "Look, I built something cool". That became the culture and led to were we are now. All this idea that they are trying to undercut the monopoly is speculation. Google still releases open models, Cohere releases command-a, Mistral releases their model too, Arcee and Thinking Machine have released models too. It's just that Chinese models have gotten good and are also leading in the open weight category and USA has a very strong paranoia of Chinese models hence this talk. Plus every time they release something, some of the labs cry, "they copied us, distillation, the bad guys have done it again!"

So what we are seeing is nothing more but geeks doing geeky stuff, the only one that has serious demonstrated that there are possibly be had is Anthropic and the Chinese labs are beginning to copying them in terms of user plans, coding tools, etc. They are giving away the model to show it's great. Once they have the compute to serve the world and they have a model just as good or better than top model, they will go close to keep all the profit.

New comment by segmondy in "OpenAI and Hugging Face address security incident during model evaluation"

segmondy — Tue, 21 Jul 2026 21:25:18 +0000

Jailbroken, all LLM models can be broken. ALL.

New comment by segmondy in "The Kimi K3 Moment"

segmondy — Sun, 19 Jul 2026 18:56:53 +0000

how do I find someone to use their number when I'm not in china and don't know anyone in china?

New comment by segmondy in "Qwen 3.8"

segmondy — Sun, 19 Jul 2026 18:17:04 +0000

let's see your grammatical correct tweet in chinese.

New comment by segmondy in "Harness Engineering"

segmondy — Sun, 19 Jul 2026 04:47:56 +0000

I don't even know how to feel about this after glancing through the repo.

New comment by segmondy in "Hardcore IndieWeb: Run your own website 100% independently for only $0.01/day"

segmondy — Sat, 18 Jul 2026 23:23:10 +0000

same here, except a 5 yrs old raspberry pi hanging off a power adapter on the floor.

New comment by segmondy in "The Kimi K3 Moment"

segmondy — Sat, 18 Jul 2026 21:41:31 +0000

how do we get chinese phone numbers?

New comment by segmondy in "Kimi K3, and what we can still learn from the pelican benchmark"

segmondy — Sat, 18 Jul 2026 00:34:27 +0000

There's a bias in the direction all things face. You can ask these models to generate a thing animal, car etc and you will notice that 90% of them will converge towards the same sort of results. If you ask for something rotating, 90% of them will rotate right and a few odd ones will rotate left.

New comment by segmondy in "Kimi K3, and what we can still learn from the pelican benchmark"

segmondy — Sat, 18 Jul 2026 00:32:42 +0000

It's incredible you can't reason to see if pelican on a bike is a thing. It's not! This has been discussed to death. You can ask any model to generate anything. Generate an SVG of earthworm and a robin boxing. Guess what? The smarter the model the better the image, doesn't matter if it's a vision model or not. I rolled my eyes at this eval when I first saw it, then I tried various ridiculous things and noticed a very strong correlation. Things that are absolutely not in the training set.

New comment by segmondy in "Flock CEO Apologizes for Calling Activists 'Terrorists'"

segmondy — Fri, 17 Jul 2026 21:09:56 +0000

Flock CEO doesn't give a fuck about that, they have been advised by public relations to manage their image since public opinion is against them. If they really cared, they will shut the entire thing down, but you know.

"Fuck you, pay me! If not me, someone else will!, profit must be had, your privacy be damned."

New comment by segmondy in "Frame – Linux X server in Assembly"

segmondy — Fri, 17 Jul 2026 21:04:07 +0000

With the crazy lack of supply for hardware and ridiculous prices, seems we are going to have to start squeezing more juice out of our existing hardware. What I would love to see is how well this runs on an ancient system. Will this for instance run on an old PI and make it snappy?

New comment by segmondy in "AWS: Inaccurate Estimated Billing Data – $1.7 billion"

segmondy — Fri, 17 Jul 2026 14:57:58 +0000

vibe coding for the win.

New comment by segmondy in "800 Canadian wildfires burning as air quality alerts extend to US"

segmondy — Thu, 16 Jul 2026 20:27:14 +0000

700 miles away and we have the worse air quality in US. something something about everyone thinks they are safe on climate change until it shows up on your doorstep.

New comment by segmondy in "Kimi K3 is now live"

segmondy — Thu, 16 Jul 2026 18:58:19 +0000

Crap, the first open weight model that really feels out of reach when it comes to running it locally at home. :-(