remoquete

Agreed. These tests should be performed on local models.

Using prompts know to be problematic? Some sort of... Voight-Kampff test for LLMs?

Because you can't. See my previous comment. https://news.ycombinator.com/item?id=43748908

Indeed. Augmentation is the way forward.

This is nice and fun for getting some fast indications on an unknown codebase, but, as others said here and elsewhere, it doesn't replace human-made documentation.

https://passo.uno/whats-wrong-ai-generated-docs/

What's the Gemini equivalent of Claude Code and OpenAI's Codex? I've found projects like reugn/gemini-cli, but Gemini Code Assist seems limited to VS Code?

AI-or-not silliness aside (gets boring real fast): I find Gemini to be faster (it matters), more reliable, with quality of responses being at the same level as Claude's. Better integration with Google apps and devices is a plus.

You can. Then again, I'm paying.

Could be, if an AI actually wrote it.

The improvement in Gemini 2.5 is real, but I wouldn't say it's miles away from Claude 3.7. The fact that web browsing still isn't in Claude in Europe bothered me. It's many little things.

Image generation is extremely good in GPT now. Claude's edge is UX. But I doubt Google won't catch up on both fronts. It has the technology and manpower.

I was a loyal Claude user until I decided to try Gemini 2.5. "After all", I thought, "I already use a Pixel phone, so it's integrated with Android. And with Google Drive. And I can get it through my Google One subscription."

And now that I'm on it, I don't think I'm going back. Google did it again.

And CDI plans better than those in year MMXXV.

I'm wondering how this protectionist trend will impact overseas hires. Will US companies continue to hire talent remotely outside of their borders? Will they be incentivized to only hire in the US?

I'm a tech optimist, but I struggle to see a fast solution to the problems described by this article that doesn't involve ending apps and screens for good.

The most interesting content I found through HN had no pictures or illustrations.

I'm seeing more reactions like this. I wonder if it's a consequence of reading AI generated drivel and not being able to tell content apart. Or perhaps people is asking for more action and less reaction. I don't know.

I don't know. I work for a B-series startup and I'm a docs engineer.

> almost nobody has (or puts in) the time to properly edit a twenty+ pages document to make it pleasant to read

That's why technical writers and editors exist.

More common than you'd think...

Some docs are better than none, but writing less is not easy.

If functional code can be written and deployed by juniors using an LLM, then perhaps the required business logic isn't that complex after all. Ask them to write a device driver or a kernel module, or code a faster algorithm for, say, stock trading, and they'll soon see the need for closer inspection and manual coding.

The reality, though, is that for most CRUD / code scaffolding, what you need the most is good knowledge of the problem space and a solid enough foundation to ask LLMs for solutions. Because, let's not forget it, people with no coding knowledge whatsoever can't get functional stuff out of LLMs as fast as non-developers.

We need to get used to a world where augmentation means "getting rid of the boring stuff at the margins". There's no heroics in doing that stuff the old way.

No matter how brilliant the game mechanics of those games were, they wouldn't have hooked me as they did without Antonov's exceptional vision. City 17 and Dunwall Will forever live in our memory.

Comments are just one bit of the docs.

To the list I'd add:

We're destroying software when we think documentation does not matter.

I'm intrigued by the conclusion. Docs-as-code, this time turning actual documentation and requirements into code? So, specifications? Back to OpenAPI?

Back to... programming languages? :)

No it's not! :-) https://github.com/theletterf/english-lang

It's not easy. Depends on the culture where one grew up, too, and family values weigh a lot in this. But when people from many different backgrounds get together to work, praise is the kind of thing that prevent networks from falling apart.

The most important is that tech must stay at the periphery of the user's attention. Of course one cannot apply this to Candy Crush...

Amber Case's book on Calm Technology and design was a great read. Perhaps as a consequence of having studied Cognitive Science, I find this one to be the best book I’ve read on feature design — and not just for software.

It's full of easily digestible insights on attention and context, with excellent examples and clear explanations. It’s almost philosophical in its apparent simplicity.