Hacker News: kissgyorgy

New comment by kissgyorgy in "Nebula Sans"

kissgyorgy — Wed, 26 Aug 2026 15:55:00 +0000

I'm not sure we are looking at the same font, this one looks way crispier and more readable than any other font I have ever seen before.

New comment by kissgyorgy in "LFM2.5 2.6B model competitive with 4x larger models"

kissgyorgy — Tue, 11 Aug 2026 14:58:38 +0000

I don't think it's even competitive with Gemma 4 E2B

New comment by kissgyorgy in "Save GPT-5.5"

kissgyorgy — Sun, 19 Jul 2026 12:27:48 +0000

Please don't be attached to a tool emotionally, it's not healthy. Think of it like a huge autocomplete or text editor, it's not really different! It's just a tool.

New comment by kissgyorgy in "The short leash AI coding method for beating Fable"

kissgyorgy — Thu, 02 Jul 2026 21:56:17 +0000

This is probably slower than writing the code yourself. Doesn't make sense to me. Using an agent without YOLO mode is not wort it.

The way I rather do it is tightly control the output by skills written yourself, prompts, plans, etc. and have the closest possible outcome you would write yourself.

New comment by kissgyorgy in "Previewing GPT‑5.6 Sol: a next-generation model"

kissgyorgy — Fri, 26 Jun 2026 19:57:24 +0000

    we expect substantial benefit for legitimate defensive work, while meaningfully constraining prohibited offensive use.

That's literally impossible. Writing an exploit agains a known vulnerability needs the exact same knowledge that defending against the exploit of the same vulnerability.

Also just making the model better at code is just making it better to writing offensive code.

New comment by kissgyorgy in "Excessive nil pointer checks in Go"

kissgyorgy — Sun, 21 Jun 2026 10:59:11 +0000

This is exactly what LLMs are really bad at. They don't have the knowledge (and don't ask for) the invariants of the system and write defensive code at every step of the way, which is not just unnecessary, it's bad because if an unexpected state still get into the system, you will never notice and bad data will flow through and makes everything unpredictable.

New comment by kissgyorgy in "GLM-5.2 is the new leading open weights model on Artificial Analysis"

kissgyorgy — Wed, 17 Jun 2026 11:43:17 +0000

I tried it today through Openrouter and the API is atrocious. I got multiple rate limit and random errors every turn.

Somebody wrote [1]; "I am never touching Minimax or GLM again. Their APIs had constant outages and I had to restart my runs multiple times — after burning money on the runs that failed midway." and I 100% agree.

The model might be good, but if the API is so bad, it's effectively useless.

[1]: https://kasra.blog/blog/i-spent-1500-seeing-if-llms-could-ha...

Diving into the Bun Rust rewrite [video]

kissgyorgy — Wed, 03 Jun 2026 10:41:43 +0000

Article URL: https://www.youtube.com/watch?v=2AqwehBu84w

Comments URL: https://news.ycombinator.com/item?id=48382243

Points: 2

# Comments: 0

New comment by kissgyorgy in "I Cancelled Claude: Token Issues, Declining Quality, and Poor Support"

kissgyorgy — Fri, 24 Apr 2026 20:47:05 +0000

I cancelled in the minute my subscription stopped working in Pi. Not going back to the slopfest what Claude Code is.

New comment by kissgyorgy in "Making MCP cheaper via CLI"

kissgyorgy — Wed, 25 Feb 2026 23:28:54 +0000

A very good example of this is playwright-cli vs Playwright MCP: https://github.com/microsoft/playwright-cli

The biggest difference is state, but that's also kind of easy from CLI, the tool just have to store it on disk, not in process memory.

New comment by kissgyorgy in "How I use Claude Code: Separation of planning and execution"

kissgyorgy — Sun, 22 Feb 2026 14:16:11 +0000

There is not a lot of explanation WHY is this better than doing the opposite: start coding and see how it goes and how this would apply to Codex models.

I do exactly the same, I even developed my own workflows wit Pi agent, which works really well. Here is the reason:

- Claude needs a lot more steering than other models, it's too eager to do stuff and does stupid things and write terrible code without feedback.

- Claude is very good at following the plan, you can even use a much cheaper model if you have a good plan. For example I list every single file which needs edits with a short explanation.

- At the end of the plan, I have a clear picture in my head how the feature will exactly look like and I can be pretty sure the end result will be good enough (given that the model is good at following the plan).

A lot of things don't need planning at all. Simple fixes, refactoring, simple scripts, packaging, etc. Just keep it simple.

New comment by kissgyorgy in "Ask HN: Why is my Claude experience so bad? What am I doing wrong?"

kissgyorgy — Sun, 15 Feb 2026 21:56:43 +0000

You need to be very specific about what to build and how to build it, what tools to use, what architecture it should do, what libraries, frameworks it should include. You need to be a programmer to be able to do this properly and it still takes a lot of practice to get it right.

New comment by kissgyorgy in "Claude Code is being dumbed down?"

kissgyorgy — Wed, 11 Feb 2026 18:37:20 +0000

This is why I am a big fan of self-hosting, owning your data and using your own Agent. pi is a really good example. You can have your own tooling and can switch any SOTA model in a single interface. Very nice!

https://lucumr.pocoo.org/2026/1/31/pi/

New comment by kissgyorgy in "CLI agents make self-hosting on a home server easier and fun"

kissgyorgy — Mon, 12 Jan 2026 15:33:45 +0000

My non-technical friend, never learned coding, doesn't know Linux, zero sysadmi experience does this and he can do anything and doesn't even know what Clause is doing. He learned some concepts recently like Docker, SSH, but that's basically it.

OpenAI might train on responses API data

kissgyorgy — Sat, 20 Dec 2025 11:57:53 +0000

The quote from their Chief Scientist in the official documentation is quite suspicious:

     the hidden chain of thought allows us to “read the mind” of the model and understand its thought process. For example, in the future we may wish to monitor the chain of thought for signs of manipulating the user.

If they don't train on it, they are definitely reading the reasoning tokens.

https://developers.openai.com/blog/responses-api

Comments URL: https://news.ycombinator.com/item?id=46335541

Points: 1

# Comments: 0

New comment by kissgyorgy in "Django: what’s new in 6.0"

kissgyorgy — Wed, 10 Dec 2025 11:49:50 +0000

I find templates atrocious to use for component fragments like this, that's why I wrote a Python component library when I started using Django with HTMX. Order of magnitude more pleasant to use, works with _every_ Python web framework not just Django: https://compone.kissgyorgy.me/

New comment by kissgyorgy in "The "confident idiot" problem: Why AI needs hard rules, not vibe checks"

kissgyorgy — Mon, 08 Dec 2025 13:14:28 +0000

It's just simple validation with some error logging. Should be done the same way as for humans or any other input which goes into your system.

LLM provides inputs to your system like any human would, so you have to validate it. Something like pydantic or Django forms are good for this.

New comment by kissgyorgy in "After Windows Update, Password icon invisible, click where it used to be"

kissgyorgy — Tue, 02 Dec 2025 12:35:45 +0000

But hey! At least these four AI components made it in, so the important stuff is okay...

New comment by kissgyorgy in "Google Antigravity just deleted the contents of whole drive"

kissgyorgy — Mon, 01 Dec 2025 17:18:05 +0000

What's the equivalent of prepared statements when using AI agents?

New comment by kissgyorgy in "Ty – A fast Python type checker, written in Rust"

kissgyorgy — Mon, 01 Dec 2025 15:20:52 +0000

Maybe I'm wrong on this, but I rather have 1 tool everyone else is using. Cargo in Rust ecosystem works really well, everyone loves it.