Goodbye to the Hard Parts That Never Mattered

echelon

This is what every holdout needs to read. The future is exciting and breathtaking, not lamentable. Stop crying about punch cards and hanging chads. Those are some thick rose-tinted glasses.

Modified3019

If reading this smells like it’s been ai generated, that’s because it is. https://web.archive.org/web/20260606173815/https://jakegolds...

Lerc

This seems to be a fairly common opinion, as does the opinion that this sits as a counterpoint to. There has been a divergence that is quite stark. I have seen it argued that this is a reflection of ability, and that as a skill multiplier those lacking in skill are feeling left behind. I'm not certain that this is accurate. I definitely see people who have made things that impressed me in the past have been more likely to embrace AI. It may just be a measure of a type of personality. I have noticed that people who have a strong sense of possessiveness over what they create are more resistant, but those who create in order to have the world contain the thing they are making are happy to have anything that will enable them to contribute to the world.

gnarlouse

Slop detected

t-writescode

I think the style essay akin to a punchy, TED-talk style persuasive essay is dead. Whether or not this author wrote it themselves (I strongly suspect at least an AI editor), the specific style that was, for a time, very attractive has become the hallmark of AI-written prose; and, while the content doesn’t seem wrong from a particular view, it also leaves a taste of waste to me - waste that I spent that time reading something they didn’t write, full of -isms that aren’t theirs, when the thesis could have padded out far less paper. The word count feels unearned, even if the content is fine.

GMoromisato

> The part I find most exciting is not that the same ticket takes fewer hours. It is that entirely different projects now fit inside a human life. This is my favorite part too. With my own project for example, I had the agent integrate libgit into it so I could add source control to my IDE. Yes, I could have done it myself, but it would have taken me months of trying to understand the API, testing various things, and finally implementing it all. Instead, the agent did almost all the work (I still designed the architecture and how it plugged in). It took about two weeks calendar time and most of that time I was doing something else while the agent worked. The thing is, before AI agents, I wouldn't have even attempted the work. I wouldn't have been able to afford the time.

rob74

Even if it's AI generated (or AI "improved"), I'll comment on the content: > The learning loop got tighter. Ask a question. Inspect the answer. Run the code. Break it. Read the implementation. Correct the assumption. Try again. This assumes that you have the resources (time and money/AI usage budget) and the interest/motivation to do that. If you do programming as a job rather than as a hobby, the time you can spend on things like this and the budget you can spare for it will be severely limited. Also, after you have finally coaxed the LLM into generating something that fits all the requirements without breaking in new and surprising ways, has code that's not ridiculously overengineered (that's probably the real reason behind vibe coding: as long as you don't look at the horror beneath the hood, you can tell yourself everything's fine), doesn't break some convention of the codebase and has the required test coverage and other metrics, you're not only out of time for the task at hand, but also more than ready to move on to something else.

bschwindHN

This guy can't even be bothered to type his own thoughts out, but expects us to read what claude dumped out? Typing your own thoughts is just not hard at all, I don't get it.

skydhash

What I understood from TFA is that the author is complaining about formalism and bad abstractions. The latter is a real complaint, but I can’t understood viewing formalism as a burden. At it’s core computing is about taking some information, encode it, transform it, and then decode the result. The latter can be interpreted by humans or used to drive some machinery. The value of computers is that they can do encoding/transform/decoding part reliably and really quickly. But it’s up to use to specify how. One common trait I found with people that dislike formalism is that they have great reluctance to admit they’re wrong, or at least consider the possibility. And a formal system cleanly mark what is correct according to its axioms. The issue with most AI practice is that they eschew reliability and guarantees of correctness (not there hasn’t been bugs previously). The proponents can talk about their goals, but they can’t explain the projects they’re working on and reason how it should be correct. For them, it’s like seeing a chess game and deciding it’s ok if a pawn move like a knight to capture the king, because that’s the intended goal. Like “just move the piece with your hands”. Because it’s hard to devise a strategy that respect the rules. The bad play is obvious when it’s a real chess board, but imagine it’s a generic board where all pieces are little puck with a led fave for colors and types. They wouldn’t mind the pawn to change its type to knight for the illegal move and then revert to pawn once that’s done. So something like Python and Go is pretty generic. You can code various domains with it (servers, games, apps,…), but it’s up to you to really follow the rules of that domain. Using AI code, there’s a non zero chance for a bit of cheating to happens. While it “works”, it’s not correct, and there’s some use cases that are totally wrong.