I try vibe coding from time to time and it’s a mixed bag. When it works it works but then there’s suddenly just code there. It’s like when a colleague implements a feature in a project with multiple devs but since it’s just your personal project you suddenly got put in the code reviewer chair.

When I write code I reason about all the small decisions but when the agent takes over I have to figure out step by step why decisions were made.

I’m still not sold on coding agents and vibecoding but it’s getting better and more usable. The only downside is that a agent doing its thing is now seen as downtime where you could just prompt another agent and work on several projects simultaneously. If anyone ever tells you vibecoding is good because you can work in parallel RUN

~~I try vibe coding from time to time and it’s a mixed bag. When it works it works but then there’s suddenly just code there. It’s like when a colleague implements a feature in a project with multiple devs but since it’s just your personal project you suddenly got put in the code reviewer chair.

When I write code I reason about all the small decisions but when the agent takes over I have to figure out step by step why decisions were made.

I’m still not sold on coding agents and vibecoding but it’s getting better and more usable.
The only downside is that a agent doing its thing is now seen as downtime where you could just prompt another agent and work on several projects simultaneously.~~ If anyone ever tells you vibecoding is good because you can work in parallel RUN

most users are also too accepting of slop work. So the MVP of yesterday is the final product of today.

Just look at any productivity vs wage growth chart from any country. Anyone who thinks that any efficiency gain will pay off for workers is either naive or part of the grift.

the plane can do a lot of it automatically these days.

It can but nobody is out there flying CAT IIIc autoland approaches because its 'easier'.

The pilots job is specifically to prevent to finding out phase.

Just install a second autopilot that can take over if the first one fails!

I watch a lot of airplane youtube, and this just seems to be wrong. During critical phases of flight, i.e landing and descending, the pilots have a high workload even with the autopilot. They are directing the autopilot ( which isn't obvious how to do), they're communicating, navigating, etc.

Flight crews of the 20th century sometimes had crews of 3 or 4 people to handle all the tasks that are more automated these days. We don't have dedicated flight engineers or navigators any more -- the pilots do these jobs now too.

During the cruise phase of flight, yes the pilots have a lower workload because of the autopilot. But on lots of flights, there might not be much of a cruise at all -- they climb, and then descend.

I don't know how much commercial pilots are required to hand fly the plane, but it is a high workload even with the automation -- the automation just helps them do a better job.

Luddites weren’t concerned about the printing press. Their fight was over mechanized looms and knitting frames.

Their concerns were mostly about losing their jobs, as well as the absolute dog shit quality the mechanical machines produced at the time.

In 1814, printers in London did have similar conniptions over the steam press, though.

I'm the "good at exams" type, I find paper writing (or novel writing... 😢) unpleasantly tedious. If we lose the "good at structuring complex thoughts in a way that can be coherently decoded by other humans" types, the "good at exams" types will suffer. We need people who can write good papers and people who can recall diverse facts and arrange them into simple shapes quickly.

I wasn't great at exams, but I think if I re-did university today I'd still do much better because my personal learning style is best when I talk through the material with someone. Sitting in a lecture listening to someone monologue at me, reading a static textbook, those weren't great. But with LLMs every student now has full-time access to a "live tutor" with infinite patience and no sense of social risk from asking it dumb questions. It's absolutely my perfect learning environment.

Yeah I would have done much better in high school like this, though I don’t think AI will eliminate the poster projects unfortunately.

Fucking scary people are so willing to turn off their brain. When shit goes wrong, they will hold you accountable, not the AI.

I think it is what the original post is talking about - it literally says "We're doing this in software now ... Nobody's asking what happens in month 18"!

I have a little more hope for future generations. Regardless of if you are primarily using ai to speed up a traditional work flow or vibe coding, there still are more and less effective ways to get things done. I've found if I write sonething I'm familiar with like bash or Python, I "know enough" to write functions and refactor incrementally. It's a methodical process which produces code better than I can write by myself in less time.

Compare that to when I write Lisp. There it's more or less throw shit at the wall until I somewhat get what I want. It takes longer and can be next to impossible to extend or effectively refactor.

Based on that observation of myself, I think we'll find two (and probably more) kinds of programmers in the future. Ones who use it as an assistant to speed up a traditional workflow, and those who produce the slop that we are afraid of.

Perhaps to give us any amount of additional hope/comparison to now. There still have been people who have programmed now who create slop. And that comes to the way we use references like stack exchange or the Arch Wiki. Are you copy and pasting while learning what things mean and learning in the process? Or are you just stringing things together to get results at all? I don't think we would argue those references made all of us worse.

So I think it's an apt concern, but we'll have to wait to see what it's like whole sale. I was a doubter for a long time, but I've come to understand where Torvalds et al are coming from.

That's a very good point

But is he opinionated or is the LLM's training corpus opinionated? Why should I care what an autocomplete has to say about this topic?

This isn't One Hour Hot Dog. If you need a big huge fix right the hell yesterday, you fucked up last week. Possibly last year. And what's gonna prevent that sort of company-liquidating blunder is a layer of crotchety obsessives who can spend a day reading and then commit one line. 'Hit engine with hammer: $1. Knowing where to hit: $9,999.'

A system that's pretty good at guessing a hundred places to try hitting has obvious value. Especially when undo is a thing. But this attitude of 'the future is now, old man!' is almost as off-base as the folks insisting a magic coding robot won't help anybody.

or we end up with a JS-based start menu in Win11

It can’t really write maintainable code without a lot of re-prompting and corrections.

Which is a skillset being learned. One which users are presumably better at than non-users, since they're practicing it. If the chatbot-that-codes Just Worked, there'd be nothing to discuss; we could consider human coding successfully completed. The people who don't move up to proper computer science can get on with whatever they wanted software for. We only teach them all of this because they've got shit to do, and up until now, it's been really really hard. Like if beginner astronomy meant telescope engineering.

Already the key complaint has become "maintainability." That's quite a high-level criticism. When I sass specific compilers, it's for using register-juggling macros that could have been one hardware instruction... not for adding a feature from a one-sentence description, but in a way that will complicate any future refactor. Or would, if refactoring was not also something you can tell the chatbot to do.

Compiler developers are a hardened core of ultranerds. I couldn't do it above a toy level. If I needed to understand how they work, in detail, then I would be in trouble. Thankfully it is sufficient to abstract all of their work into 'C goes in ARM comes out.' Or at least to mumble about intermediate representations and distract people with Godbolt before running away. Sure, 'add a button above the widget' is more pluripotent than a K&R-standard Hello World, but even the foundation you celebrate is layer upon layer of Don't Worry About It.

AI coding right now is a blunt instrument. Those often suffice. And this one can already be used to remove layers. To obviate a fat framework. To translate to a faster language. To golf. Or it could format your hard drive if a website tells it to, IDFK. It is a higher class of footgun. We take those.