Please, be patient. Wrangling AI agents, writing and rewriting prompts, waiting for the start of another month because tokens ran out - there are so many challenges here, you cannot expect everyone to ship an app a day or something.

Thoughts: 1. Some hype-types may have been effusive about AI-assisted coding since ChatGPT, but IMO the commonly agreed paradigm shift was claude code, and especially 4.5, very very recent. 2. Anchoring biases in reaction to hype is still letting one's perspective be defined by hype. Yes the cursor post is a joke, but leading with that is a strawman. This article does not aim to take it's subject seriously, IMO. 3. While I agree the hype is currently at comical levels, the utility of the current LLMs is obvious, and reasons for "skilled" usage not being easily quantifiable are also obvious.

IE, using agents to iterate through many possible approaches, spike out migrations, etc might save a project a year of misadventures, re-designs, etc, but that productivity gain _subtracts_ the intermediate versions that _didn't_ end up being shipped.

As others have mentioned, I think yak-shaving is now way more automated. IE, If I want to take a new terminal for a spin, throw together a devtool to help me think about a specific problem better, etc, I can do it with very low friction. So "personal" productivity is way higher.

I’m not a developer by trade. I’ve screwed around with some programming classes when I was in school, and have written some widely used but highly specific scripts related to my work, but I’ve never been a capital-D developer.

In the last few months, Gemini (and I) have written for highly personal, very niche apps that are perfect for my needs, but I would never dream of releasing. Things like cataloguing and searching my departed mom‘s recipe cards, or a text message based budget tracker for my wife and I to share.

These things would never be released or available as of source or commercial applications in the way that I wanted them, and it took me less time to have them built with AI then it would have taken me to Research existing alternatives and adapt my workflow/use case to fit whatever I found.

So yeah, there are more apps but I would venture to say you’ll never see most of them…

AI does make me more productive. At least until the stage of getting my idea to the "working prototype stage". But in my personal experience, no one has been realistically able to get to the 10x level that a lot of people claim to have achieved with LLMs.

Yes, you do produce more code. But LoC produced is never a healthy metric. Reviewing the LLM generated code, polishing the result and getting it to production-level quality still very much requires a human-in-the-loop with dedicated time and effort.

On the other hand, people who vibe code and claims to be 10x productive, who produces numerous PRs with large diffs usually bog down the overall productivity of teams by requiring tenuous code reviews.

Some of us are forced to fast-track this review process so as to not slow down these "star developers" which leads to the slow erosion in overall code quality which in my opinion would more than offset the productivity gains from using the AI tools in the first place.

I AI coded an entire platform for my work. It works great for me. I also recognize that this is not something I want to make into a commercial product because it was so easy that there's just no value.

I think this might be more of an comment on software as a business than AI not coding good apps.

They're private, that's the beauty. Code is so cheap now, we can ween ourselves off massive dependency chains.

200 years ago text was much more expensive, and more people memorized sayings and poems and quotations. Now text is cheap, and we rarely quote.

Vibe coding is actually a brilliant MLM scheme: people buy tokens to generate apps that re-sell tokens (99% of those apps are AI-something).

The models are not well trained on bringing products to market.

And even “product engineers” often do not have experience going from zero to post sales support on a saas on their own.

It is a skill set of its own to make product decisions and not only release but stick with it after the thing is not immediately successful.

The ability to get some other idea going quickly with AI actually works against the habits needed to tough through the valley(s).

This is going to cause people to react, but I think those of us that truly love opensource don't push AI generated code upstream because we know it's just not ready for use beyond agentic use. It's just not robust for alot of use common use cases because the code produces things that are hyper hardcoded by default, and the bugs are so basic, i doubt any developer that actually cared would push something so shamefully sloppy upstream with their name on it.

The tools for generating AI code aren't yet capable of producing code that is decent enough for general purpose use cases, with good robust tests, and clean and quality.

Where are they? Well they aren't being uploaded to PyPI. 90% of the "AI apps" one-off scripts that get used by exactly one person and thrown away. The rest are too proprietary, too personal, or too weird to share.