entropyneur

all the vibe coded stuff is still slop (where’s all this new magical software that the productivity improvements should imply?)

Part-time vibe coder here. As far as I can tell it's no longer "slop" in the sense that I am no longer hitting the state where AI can't maintain what it has written or can't meet the requirements. But shipping is as hard as ever. Projects get more ambitious, scope creeps, nuances keep being discovered as you work on a piece of software, etc. Right now I feel that the absence of new software explosion is best explained by the fact that the part we have automated turned out to be relatively small.

Sorry, I understand scraping is a problem, but talking about open Internet while simultaneously complaining you can no longer discriminate datacenter IPs like you used to is hypocrisy.

I use a datacenter-based IPv6 address because my local ISPs don't offer v6 connectivity and the Internet is already broken for me. And generally the entire idea of a "residential" IP address smells.

Around last fall it started chronically lagging in SOTA models support. Honestly, not sure why a code update is even needed for that. Eventually it'd get patched but it just wasn't quick enough for me. Other development seems to have ceased completely. At one point I tried to investigate what's going on and it looked like the community had moved on. There was a fork that promised a liberal approach to accepting changes and it looked broken beyond repair because of that.

This reminds me of the workflow I had a year ago. Miss Aider so much. Are there any good open source agents right now? Might be a good time to try one soon as Fable switches to token-based billing, which Code is designed to maximize.

Yep.

Whew. Crashed before I sunk my day there.

It's disappearing. Even if models stop evolving tomorrow, there's still enough potential in the harness improvement to reach the point where anything 99% of us know about software engineering is useless. How many humans will be involved in creating software after the dust settles is anybody's guess, but I wouldn't bet on it being anywhere near the current level.

How about fixing whatever the hell prevents competitive private LLM vendors from appearing in Europe?

Sounds like calling those model "open source" did its wicked job. You can not take an open weight model and build a next-generation model using that as a foundation. Once those companies decide it's no longer in their interest to release new open weights everything you've created this way becomes a pile of rapidly deprecating legacy.

The new tricks seem to be "ultrathink" and "you are a rock star software architect". That's not the stuff careers are made of.

I don't care as much about LLMs deprecating my accumulated knowledge. After a quarter of a century in the industry I have plenty of stuff in my head of purely sentimental value.

But for me, the real catastrophe is that they took away all my motivation to learn which was the main work driver for me. Anything I can learn now, the models probably already know or will learn soon enough. Steering LLMs isn't anywhere near being a "deep" skill I'm used to having and it too is being eaten by the agentic tools faster than we learn it. The "make it good" button is coming. And I hate it.

Personally, I don't know a place more hyped up about AI than HN. It turned from my daily dose of tech excitement into a daily dose of tech anxiety.

As for your argument, there's no such thing as elegance. Code "elegance" is mainly maintainability (and, to a smaller degree, some other aspects like security, performance, etc.). The importance of maintainability greatly varies between projects, industries and individual subjective viewpoints, resulting in the diversity of attitudes to AI-assisted coding. That, of course, assumes that AI cannot match humans in maintainability. Which seems to be the case to me right now. But it also seems that the gap is closing, not as much through AI writing "better" code, but mainly through it being increasingly capable of maintaining "bad" code.

I've recently found myself unable to finish articles that take more than a paragraph to announce their point. But starting with "Your life’s goal should be" is a level of boldness I wasn't prepared for.

I had human teachers who did that in middle/high school. Took me many years to pick out all the hallucinated bits of "knowledge". I don't think the current models are any less reliable that what we currently have on average.

I've tried switching from JetBrains IDEs just a few days ago. The speed and memory footprint are very impressive. I ended up badly missing refactorings and some other features and configuring a debugging session looked like something that needs more time than I had on my hands. So went back for now. I hope they add more IDE features eventually. There's not much a pure text editor can offer over Emacs after all. But this announcement sounds like they are prioritizing agents integration - the same thing that seemingly made JetBrains drop the ball on their core advantages.

Satoshi is a paper billionaire - he can't use a small fraction of his "wealth" to hire proper security. Simultaneously his "assets" are much more attractive to criminals. Imagine holding a regular billionaire hostage and demanding they give you a billion dollars. They'd probably have to sell 1B worth of stock, then convert it to cash (or crypto), etc. all of that requiring multiple interactions with different people and institutions.

The real winner in this war is Israel. Iran's military might is now a shadow of its former self while all the costs have been paid by someone else: American taxpayers, gas consumers around the world, Arab states. Even the political costs are on Trump.

I am working on a P2P VPN app that lets you route your traffic through a friend's Internet connection easily. It has a few distinct uses, but right now I am testing whether emigrants from authoritarian countries will be interested in providing censorship-free connectivity to their family/friends at home.

It's called Spora: https://spora.to/

The current MVP is about 90% vibe-coded while being a fairly sophisticated piece of software.

It's not the extension developer who should decide this, but the browser user.

Not a mental health crisis like the guy in TFA had, but I've definitely experienced states I would characterize as overexcitement while calibrating my expectations of these new tools to their abilities.

It was close enough for me and I do acknowledge the cruelty and abstain from many kinds of meat. I was super excited when I tried it first. But after about a year of being part of my regular diet it started being disgusting unfortunately. Now I can only eat it once a in a while.

It may indeed be the case that the candidate promised one thing and the voters acting irrationally (or correctly assuming he's a liar) voted with an expectation of him doing the exact opposite. The GP, however didn't say anything about voting. He was talking specifically about the mismatch between campaign promises and actions taken once in office.

I am working on a P2P VPN app that lets you use a friend abroad as your VPN provider with no special setup: https://spora.to

It's mainly for censorship evasion (should be much harder to block than the regular centralized VPNs), but also for expats to access geo-blocked domestic services.

It's at the MVP stage and honestly it evoked much less interest in people than I hoped it would, but I'm still going on despite my better judgement.

Same at 42. I've been making software for 30 years and the gap between what I can envision and what I can code in a single day is so huge that it takes all the steam out of me. With agentic coding I can move at a pace that feels right again.

If only we had some technology that would relieve us of the need to share IP addresses among multiple servers, this might have been unnecessary.

Comparatively few people live under worse authoritarianism than the one in China. Definitely not enough to form a talent pool that would make any dent in whatever China already has. Especially when you factor in education quality.

Why would anyone want to become a Chinese citizen? How's everyone discussing linguistics while completely ignoring the authoritarian elephant in the room?

Previous discussion: https://www.reuters.com/world/us-plans-online-portal-bypass-...

Weird title, but worthy of discussion. From the little info available so far this appears to be little more than political posturing. If you want to fight censorship, an "online portal" to access all the censored content is the wrongest possible way to go about it. But we'll see.

Why is it depressing? Personally, unless the alternative is literally starving, I wouldn't want to do a job that a robot could do instead just so that I could be kept busy. That sounds like an insult to human dignity tbh.

Emacs is an interesting analogy. I've switched from IDEs to Emacs at some point in my career, and inertia obviously wasn't the reason. Then another 15 years later I went back to using IDEs (inspired by Carmack's interview). 2 years in I realized it destroyed my ability to generate and maintain mental maps of the codebase and generally remember things about it, although I think it's still a net gain so I stick with it. Agentic coding poses the exact same problem and the exact same tradeoff. And I think the jury is still out on whether it's worth it. At the very minimum you need to take proactive measures least you end up with a codebase no one can maintain (other than maybe future AGI).