I agree with the authors observations here. I think rather than it being purely language related, there's a link to the practice of 'rubber ducking', where when you start to explain your problem to someone else it forces you to step through the problem as you start to explain the context, the steps you've tried and where you're stuck. I think LLMs can be that other person for us sometimes, except that other person has a great broad range of expertise.

As AI coding assistants have become increasingly capable, I've developed a structured approach to integrating them into my development workflow. This post details how I've organised my codebase to work effectively with multiple AI agents - GitHub Copilot, Claude and others, while maintaining a single source of truth for instructions across multiple projects that share a similar tech stack and structure.

https://g9n.com - a place for mini projects, apps and blog posts that don't fit into any 'work' category!

The thing about Claude code, is that it's usually used in version controlled directories. If Claude f**s up badly, I can revert to a previous git commit. If it runs amock on my office documents, I'm going to have a harder time recovering those.

It's official, vibe coding is legit.

No it’s not, there are so many issues yet to be solved. Privacy, responsibility, etc. It literally says nothing for prototyping, we have all done it.

We had all those problems before vibe coding was a thing too.

No tests, mypy type hinting or doc strings -- just pure vibes

Depends on stage. Codex/Claude-generated code have hinting and doc strings, and will gladly add tests for every change you're doing. (That's why vibe coded projects have a gazillion tests.)

I wa referring to linus' script in the repo

it's always been.

You mean you want it to be.

I use it daily and it just works. If you haven't found your agentic workflow or you don't prompt it well, that's not an ecosystem problem.

That is my point it’s a you thing. Maybe your QC are lower than mine? I don’t know. No one else on my team uses it for that reason alone.

I don't know a single person who isn't just coasting at mag7 who isn't using genAI, and I know most of them are expected to do so.

My product NumeroMoney (https://www.numeromoney.com) is the first I've built that makes over $500, and it's grown surprisingly quickly. I built it because I needed something simpler that the existing solutions I could find for understanding our families spending (YNAB etc were geared too much towards budgeting). It helps users to import and categorize bank statement transactions in a way that makes it really easy to make decisions about household spending.

Just out of interest, is this a US based product? I'm in Australia and most of the major banks have this as a built-in capability in their apps.

I'm working on https://www.numeromoney.com/pricing I don't even have the home page put together yet so marketing is still on the starting blocks! It's web app for helping to understand how you spend your money. I'm keeping it as simple as possible while trying to surface clear information about a persons spending. It came out of personal need (young families are expensive, it turns out!), and the existing products out there - YNAB etc were just too focused on budgeting. I just wanted to know where my money goes so I can focus on where I'm not spending it well.

I'd been finding a lot of the existing calculator tools out there to be not quite what I needed, so I made this one. I think the options are the right ones for the most common contractor accounting setups in the UK.

I agree it needs more images. On desktop there is space to show some encouraging text changes as you go. The unfathomable amount of scrolling required is sort of the point.

They are Gemini generated images. The original idea was to get my kids to draw pictures for each change, but their enthusiasm was not quite what I envisaged! Wanted to get it out there so AI to the rescue on that one. I agree it does cheapen the experience a bit :)