I’ve been using these for constrained, boring development tasks since they first came out. “Pro” versions too. Like converting code from one language to another, or adding small features to existing code bases. Things I don’t really want to bother taking weeks to learn, when I know I’ll only be doing them once. They work fine if you take baby steps, make sure you do functional/integrated testing as you go (don’t trust their unit tests–they’re worthless), and review EVERYTHING generated. Also, make sure you have a good, working repo version you can always revert to.
Another good use is for starting boilerplate scaffolding (like, a web server with a login page, a basic web UI, or REST APIs). But the minute you go high-level, they just shit the bed.
The key point in that article is the “90%” one (in my experience it’s more like 75%). Taking a project from POC/tire-kicking/prototype to production is HARD. All the shortcuts you took to get to the end fast have to be re-done. Sometimes, you have to re-architect the whole thing to scale up to multiple users vs just a couple. There’s security, and realtime monitoring, and maybe compliance/regulatory things to worry about. That’s where these tools offer no help (or worse, hallucinate bad help).
Ultimately, there’s no substitute for battle-tested, scar-tissued, human experience.
Do not use them.
Saved you a click.I think the most interesting, and also concerning, point is the eighth point, that people may become busier than ever.
After guiding way too many hobby projects through Claude Code over the past two months, I’m starting to think that most people won’t become unemployed due to AI—they will become busier than ever. Power tools allow more work to be done in less time, and the economy will demand more productivity to match.
Consider the advent of the steam shovel, which allowed humans to dig holes faster than a team using hand shovels. It made existing projects faster and new projects possible. But think about the human operator of the steam shovel. Suddenly, we had a tireless tool that could work 24 hours a day if fueled up and maintained properly, while the human piloting it would need to eat, sleep, and rest.
In fact, we may end up needing new protections for human knowledge workers using these tireless information engines to implement their ideas, much as unions rose as a response to industrial production lines over 100 years ago. Humans need rest, even when machines don’t.
This does sound very much like what Cory Doctorow refers to as a reverse-centaur, where the developer’s responsibility becomes overseeing the AI tool.
This is exactly why I laughed out loud, incredulously, at Dell’s “AI powered laptop” commercial that promises you will “free up so much time for the things you love” by using AI.
From washing machines to robot assembly, we’re still buying that old lie??
The tools improve, the expectations increase, the wage stays the same.
I’m confused about the washing machine part. Do they not save time?
Related: Jevons Paradox
Good point, I wonder if elasticity of demand will be affected by the AI bubble bursting (I expect there will be a recession and demand will get less elastic, bit I’m no economist)
deleted by creator
Will #7 shock me?
Number 5 made me rethink my life, get an anchor tattoo and turn to the sea.
On a container ship?
First on a big container ship, then on a midsize cruise ship, then on a riverboat, then on a two-person inflatable paddle boat. When I finally clung to a pool noodle, I realized it was time to get rid of the tattoo and return to the shore.
Since 1990, I’ve programmed in BASIC, C, Visual Basic, PHP, ASP, Perl, Python, Ruby, MUSHcode, and some others. I am not an expert in any of these languages—I learned just enough to get the job done. I have developed my own hobby games over the years using BASIC, Torque Game Engine, and Godot,
I think this is where AI unquestionably shines: switching languages/projects frequently, on personal projects.
so I have some idea of what makes a good architecture for a modular program that can be expanded over time.
But I actually draw the opposite conclusion. The architecture and maintainability needs are where AI is pretty poor, and they’re vastly different and more important in a 100-1000 person 10 year production system.
Agreed. As an ex-technical lead and co-architect i also agree that what ai does is often very poor architectural design and i wouldnt want it to touch that, ever.
This guy just vibe coded a bunch of slop, based on high quality training data (everyone’s code on GitHub, including probably lots of unity projects, godot, etc). It’s sort of disgusting to me.
It’s starting to slowly show up. I think I’ve seen quite a lot new tools and projects pop up, from World of Warcraft addons, CIs, through “game engine based on Tesla’s Aether theory” to secure loginless messengers.
I remember few months ago that the state was “If vibe coding is so good, where are the AI coded projects?”, and I’m starting to slowly feel, at least anecdotally and in the past few weeks, that they are slowly starting to surface.
As a DJ, AI music made it extremely difficult for me to build sets, since I really don’t want to support AI music. If FOSS vibe-coded apps start popping out just like AI music did, it’s going to suck - especially as someone who likes to look for new tools and cool software often, mostly around cybersecurity. Vetting tools as safe to use is already pretty difficult in that scenario.
Thankfully, most of them will just have agents.md or ./claude, so I know I can disregard them outright. Unfortunately, seems like Bitwarden is one of those :(
FUCK goddamnit Bitwarden
We’re all just beta testers.
Paying beta testers.
I’m not paying.
