Hacker News: swalsh

New comment by swalsh in "MAI-Thinking-1"

swalsh — Tue, 02 Jun 2026 21:45:25 +0000

I'd assume it's not up to par with Qwen-3.5 then, which has been distilling Claude, and the quality of the model is probably a direct result of that.

New comment by swalsh in "Claude Opus 4.8"

swalsh — Thu, 28 May 2026 22:17:20 +0000

Its amazing how quickly ive just become accustomed to being a max subscriber. I dont think I could go back to pro.

New comment by swalsh in "I think Anthropic and OpenAI have found product-market fit"

swalsh — Wed, 27 May 2026 20:59:55 +0000

Open source models, especially qwen are pretty dang good. But its not opus 4.6, the evals dont tell the full story. I question the assumption open source models are 3-6 months out.

New comment by swalsh in "Higher usage limits for Claude and a compute deal with SpaceX"

swalsh — Wed, 06 May 2026 17:25:51 +0000

Fuck, I loved grok 4.1, it was a really capable model for the money.

I'd run agents consuming hundreds of millions of tokens for less than a hundred dollars.

New comment by swalsh in "Higher usage limits for Claude and a compute deal with SpaceX"

swalsh — Wed, 06 May 2026 17:24:25 +0000

Billions in revenue just before your IPO isn't a bad deal either.

New comment by swalsh in "Higher usage limits for Claude and a compute deal with SpaceX"

swalsh — Wed, 06 May 2026 17:22:55 +0000

There's always money in the giggawatt datacenter

New comment by swalsh in "Higher usage limits for Claude and a compute deal with SpaceX"

swalsh — Wed, 06 May 2026 17:21:37 +0000

Models are a commodity, let's say Elon actually figures out building datacenters in space, or maybe he continues to be the leader of building earth based datacenters. Probably better business to not have yourself as your only customer. Dogfood, and open it to all.

New comment by swalsh in "Higher usage limits for Claude and a compute deal with SpaceX"

swalsh — Wed, 06 May 2026 17:15:59 +0000

I have never come close to my weekly limit, but have hit my hourly limit frequently.

New comment by swalsh in "He asked AI to count carbs 27000 times. It couldn't give the same answer twice"

swalsh — Wed, 29 Apr 2026 13:08:01 +0000

It amazes me how much people try to build AI systems relying on nothing more than the models knowledge. I suspect a great deal of "failed" AI experiments we keep reading are people just not having any idea how to use AI at what its good at.

New comment by swalsh in "Qwen3.6-27B: Flagship-Level Coding in a 27B Dense Model"

swalsh — Wed, 22 Apr 2026 22:48:32 +0000

Try running with Open Code. It works quite well.

New comment by swalsh in "ChatGPT Images 2.0"

swalsh — Tue, 21 Apr 2026 22:25:52 +0000

Been using the model for a few hours now. I'm actually reall impressed with it. This is the first time i've found value in an image model for stuff I actually do. I've been using it to build powerpoint slides, and mockups. It's CRAZY good at that.

New comment by swalsh in "System Card: Claude Mythos Preview [pdf]"

swalsh — Tue, 07 Apr 2026 20:58:02 +0000

My understanding is GPT 6 works via synaptic space reasoning... which I find terrifying. I hope if true, OpenAI does some safety testing on that, beyond what they normally do.

New comment by swalsh in "Show HN: GovAuctions lets you browse government auctions at once"

swalsh — Mon, 06 Apr 2026 20:52:02 +0000

I had no idea one could buy a Blackhawk for $1.5M

New comment by swalsh in "Google releases Gemma 4 open models"

swalsh — Fri, 03 Apr 2026 00:00:41 +0000

Its coding to coding. I could care less how the model is architected, i only care how it performs in a real world scenario.

New comment by swalsh in "Google releases Gemma 4 open models"

swalsh — Thu, 02 Apr 2026 19:38:58 +0000

I gave the same prompt (a small rust project that's not easy, but not overly sophisticated) to both Gemma-4 26b and Qwen 3.5 27b via OpenCode. Qwen 3.5 ran for a bit over an hour before I killed it, Gemma 4 ran for about 20 minutes before it gave up. Lots of failed tool calls.

I asked codex to write a summary about both code bases.

"Dev 1" Qwen 3.5

"Dev 2" Gemma 4

Dev 1 is the stronger engineer overall. They showed better architectural judgment, stronger completeness, and better maintainability instincts. The weakness is execution rigor: they built more, but didn’t verify enough, so important parts don’t actually hold up cleanly.

Dev 2 looks more like an early-stage prototyper. The strength is speed to a rough first pass, but the implementation is much less complete, less polished, and less dependable. The main weakness is lack of finish and technical rigor.

If I were choosing between them as developers, I’d take Dev 1 without much hesitation.

Looking at the code myself, i'd agree with codex.

New comment by swalsh in "Qwen3.6-Plus: Towards real world agents"

swalsh — Thu, 02 Apr 2026 19:12:05 +0000

Try using Grok 4.1 reasoning. It's crazy cheap, and really it's not that bad.

New comment by swalsh in "Nvidia NemoClaw"

swalsh — Wed, 18 Mar 2026 17:26:13 +0000

Neurons that fire together, wire together. Your brain optimizes for your environment over time. As we get older, our brains are running in a more optimized way than when we're younger. That's why older hunters are more effective than younger hunters. They're finely tuned for their environment. It's an evolutionary advantage. But it also means that they're not firing in "novel" ways as much as the "kids". "kids" are more creative I think because their brains are still adopting, exploring novelty, neuron connections aren't as deeply tied together yet.

This is also maybe one of the biggest pitfalls as our society get's "older" with more old people, and less "kids". We need kids to force us to do things differently.

New comment by swalsh in "OpenRocket"

swalsh — Wed, 18 Mar 2026 17:13:43 +0000

Oh i've been looking for a project for my 11 year old... he's a very project oriented learner, which schools don't seem to do anymore.

New comment by swalsh in "Codegen is not productivity"

swalsh — Sun, 15 Mar 2026 17:17:49 +0000

Speak for yourself, I have never thrown away code at this rate in my entire career. I couldn't keep up this pace without AI codegen.

New comment by swalsh in "Show HN: Rust compiler in PHP emitting x86-64 executables"

swalsh — Wed, 04 Mar 2026 09:15:19 +0000

I bet claude was hyping this guy up as he was building it. "Absolutely, a rust compiler written in PHP is a great idea!"