kissgyorgy

6 ms·

31.
by kissgyorgy 1y ago
It's MIT licensed, can be easily picked up by someone else.
32.
by kissgyorgy 1y ago
If you don't want to upgrade and follow model development so much, I would just pay one provider and stick with them. This model worth knowing about, because it's 3x cheaper and 2x faster than the previous Claude model.
33.
by kissgyorgy 1y ago
I think clean code is more important than ever. LLMs can work better with good code (no surprise), and they are trained on so much shit code they produce garbage in terms of clean code. They also don't have a good taste or deeper archi
34.
by kissgyorgy 1y ago
If you want to keep up with AI progress and model updates, simonw is the man to follow!
35.
by kissgyorgy 1y ago
The huge advantage of SQLite is not that it's on the same machine, but it's that is it in process which makes deployment and everything else just simpler.
36.
by kissgyorgy 1y ago
I think it depends on the domain. For example, GPT-5 is better for frontend, React code, but struggles with niche things like Nix. Claude's UI designs are not as pretty as GPT-5's.
37.
by kissgyorgy 1y ago
One can only hope this would bankrupt companies doing this and other companies would learn not to push AI into every fucking thing.
38.
by kissgyorgy 1y ago
What I did recently when developing a TUI was that I put the state in a dict, start the app in an infinite loop and whenever it quit, reload module, keep the state and instantiate the class with that state again. Something like this:
39.
by kissgyorgy 1y ago
> Firstly, what is AGI? AGI is the biggest succesful scam in human history Sam Altman came up with to get the insane investment and hype they are getting. They are intentionally not defined what it is and when will be achieved, making it
40.
by kissgyorgy 1y ago
It's so much easier to build a mental model of a code base with LLMs. You just ask specific questions of a subsystem and they show files, code snippets, point out the idea, etc. I just recently took the time to understood how the GIL w
41.
by kissgyorgy 1y ago
ahh, sorry, different article :(
42.
by kissgyorgy 1y ago
Not even close. An agentic tool can be fully autonomous, an IDE like Cursor is, well it's "just" an editor. Quite the opposite. Sure it does some heavy lifting too, but still the user writes the code. They start to implement
43.
by kissgyorgy 1y ago
This is explained in 3.2 How to design good tools? This saves the LLM from having to do multiple low level clicking and typing and keeps it on track. Help the poor model out, will ya!?
44.
by kissgyorgy 1y ago
putting the review into git notes might have worked better. It's not attached to tje lines directly, but the commit and it can stay as part of the repo
45.
by kissgyorgy 1y ago
Not at all. Good documentation for humans are working well for models too, but they need so much more details and context to be reliable than humans that it needs a different style of description. This needs to contain things that you would
46.
by kissgyorgy 1y ago
I was thinking about this too, but the problem is that different models need to be prompted differently for better performance. Claude is the best model for tool calling, you might need to prompt less reliable models differently. Prompt eng
47.
Reverse-Engineered Claude Code: Learn These Agent Tricks [video](youtube.com)
6 points by kissgyorgy 1y ago|0 comments
48.
by kissgyorgy 1y ago
Claude can already do this: https://docs.anthropic.com/en/docs/claude-code/github-action...
49.
by kissgyorgy 1y ago
Context engineering is extremely hard,you won't just "get it" overnight. The other thing, it's just another code generation tool, treat it accordingly
50.
by kissgyorgy 1y ago
The original title is "Objects should shut the fuck up". I don't like unnecessary cursing either, but it is emphasizing his frustration in this case and cursing objects, not people. Renaming the title is just losing informati
51.
by kissgyorgy 1y ago
Not sure I would consider a user copy-pasting an URL being a bot. Should curl be considered a bot too? What's the difference?
52.
by kissgyorgy 1y ago
This implementation is unnecessary complicated. For the step update, you can use Out Of Band update: https://htmx.org/attributes/hx-swap-oob/ which works in a way that you can send multiple HTML fragments which ca
53.
by kissgyorgy 1y ago
[flagged]
54.
by kissgyorgy 1y ago
This sucks out the joy of Open Source development, which is freedom to do anything, change anything and not responsible for anyone else. At the point you require payments, users can have expectations and requests, which turns your project i
55.
by kissgyorgy 1y ago
I have a rule for Claude Code to disallow deleting anyhting: "deny": [ "Bash(rm:*)" ]
56.
by kissgyorgy 1y ago
I don't understand why people only think in extremes. You don't have to give up anything you did before at all. LLMs just here to increase your productivity, but cranking out unreviewed code where you don't want to do that is
57.
by kissgyorgy 1y ago
It is composable with all decades old Linux CLI tools, which you simply can't do with an IDE. It also doesn't prevent you from using an IDE at all, but still fits for people with text editors like Vim who doesn't want to use
58.
by kissgyorgy 1y ago
AGI is a scam. I'm pretty sure every big name in AI knows it's nowhere near and LLMs won't get us there. It's just marketing helping Sam and alike to get those billions and the hype alive.
59.
by kissgyorgy 1y ago
Why don't use your own personal access token?
60.
by kissgyorgy 1y ago
The only thing crazier than asking for a personal access token is that people probably do it.

More ›