Hacker News: mromanuk

New comment by mromanuk in "FLUX 3 Image"

mromanuk — Fri, 02 Oct 2026 19:11:36 +0000

For a moment I was confused that this was a release of a new stable diffusion. What happened with Stable Diffusion?

New comment by mromanuk in "Bend 2 and the Vibe-Coding Trap"

mromanuk — Fri, 18 Sep 2026 15:53:44 +0000

> The problem is that vibe coding makes it possible to build a substantial solution before learning enough about the problem to recognise that a much better solution exists

As a software developer we should fear chasing "better" solutions, that path always lead to procrastination, "kitchen sink" and probably not what users wants. "Good enough" should suffice in most cases. Sure if you are building a super mega critical software to land a plane or something like that, is different. But every day software, shouldn't be treated like "carved in stone" and stuff that should last a 1000 years. Code can be cheap now (calling it "vibe coding" doesn't help).

You can create or modify something quite fast now. Better to focus on testing, documentation, making sure that software will do what is expected.

New comment by mromanuk in "Muse Spark 1.3"

mromanuk — Wed, 02 Sep 2026 20:13:21 +0000

I didn't like 1.2, It make some mistakes in a web app, so I quickly went back to Claude, Kimi K3 or Deepseek V4. Hope this one can clear agentic development, because Muse Spark models are fast and cheap.

New comment by mromanuk in "GLM-5.3-Flash"

mromanuk — Wed, 26 Aug 2026 19:48:05 +0000

You can download the weights, and run in your own hardware and avoid all that.

New comment by mromanuk in "Don't paste the AI, please"

mromanuk — Thu, 20 Aug 2026 13:15:44 +0000

> The person on the other side has the same tools you do. If they wanted the generic answer, they would have gotten it in four seconds. They asked you because they wanted your take on it... Your context, your taste, your judgement.

I beg to differ. Most of my life, as “the computer guy” — basically since I was a kid — I’ve always been surprised by how lazy people can be, and how they often prefer to ask “the computer guy” or some other human trusted source rather than work out the answer themselves.

There was even a website called “Let Me Google That for You” back when Google was a trustworthy source of answers, precisely to point out how easy it was to find the answer yourself. Obviously, that didn’t stop people from asking, and I don’t think LLMs will either.

People will continue to turn to “the computer guy” — aka you. And if your judgement is that copying an LLM response verbatim is the best answer, then yes, you’re fulfilling your role perfectly well.

New comment by mromanuk in "AGI-64 Brings Sierra Adventures to the Commodore 64"

mromanuk — Mon, 17 Aug 2026 12:44:28 +0000

I know what is AGI, I read the queues Adventures, Sierra, should be obvious where this lead. But my mind went to "oh, thanks to LLM coding agents, they massively ported every Sierra Adventure to C64".

Show HN: A declarative, SwiftUI toolkit that renders to

mromanuk — Wed, 12 Aug 2026 10:53:02 +0000

Article URL: https://pixdeo.github.io/weavekit/

Comments URL: https://news.ycombinator.com/item?id=49270419

Points: 3

# Comments: 0

New comment by mromanuk in "Declarative Canvas UI for the browser, inspired by Swift"

mromanuk — Tue, 11 Aug 2026 18:49:46 +0000

Thank you! Yes, SwiftUI is really nice to write, and even better to read.

I'm thinking of this as a Canvas toolkit for building apps for desktop, mobile, and the web. Let's see how this experiment goes.

Declarative Canvas UI for the browser, inspired by Swift

mromanuk — Tue, 11 Aug 2026 18:27:42 +0000

Article URL: https://romanuk.org/declarative-canvas-ui-for-the-browser-inspired-by-swift/

Comments URL: https://news.ycombinator.com/item?id=49262465

Points: 4

# Comments: 2

Show HN: WeaveKit – A declarative, SwiftUI like toolkit that renders to canvas

mromanuk — Tue, 11 Aug 2026 16:11:13 +0000

Article URL: https://github.com/pixdeo/weavekit

Comments URL: https://news.ycombinator.com/item?id=49260486

Points: 2

# Comments: 0

New comment by mromanuk in "Prime Agent: A self-improving RLM agent"

mromanuk — Thu, 06 Aug 2026 15:28:02 +0000

They achieved 95% in ARC-AGI-3. [0]

0: https://x.com/PrimeIntellect/status/2085087000764568010

Prime Agent a general-purpose coding harness. On ARC-AGI-3 scored 95.5%

mromanuk — Thu, 06 Aug 2026 00:18:33 +0000

Article URL: https://x.com/PrimeIntellect/thread/2085087002769379520

Comments URL: https://news.ycombinator.com/item?id=49190818

Points: 2

# Comments: 0

New comment by mromanuk in "Kimi-K3 Releases on HuggingFace 7/27"

mromanuk — Mon, 27 Jul 2026 12:47:07 +0000

Are they going to release Kimi K3.1? I’m eager to test it. According to rumors on X, it could outperform Fable. Could it be the first Chinese open-weight model to become the leading frontier model?

New comment by mromanuk in "Testing Gemini 3.5 Flash Lite for human detection in home surveillance"

mromanuk — Mon, 27 Jul 2026 12:44:14 +0000

I should’ve posted some of the clips, especially the ones where the model mistakes the dog for a person. It mostly happens when the dog is stretching out on the floor. That's because the camera is positioned higher up, looking down at an angle. I’ll update the post soon.