loktarogar

oh yeah - in this case the result is that the code was smaller and they were happy with that. correctness was not a criteria

If you replaced "AI" with "a developer" in this article, it might make more sense why it's being dunked on. It would be an article about someone telling someone to do something, and then when the work was done they were happy with the result. It's just a bit of a nothing all round

I mean. they're selling it to a competitor so it's not really an advantage

yes, but it's only advantage if one is compute-strained and the other isn't. if they both have lots then there's no advantage. if one doesn't fully utilise their compute then it's not an advantage either

Would you accuse someone of murder with 26% matching evidence?

The Director of Photography on the movie is also an Australian!

I'm building something that has to share a pool of phone numbers for SMS between many businesses with many clients and the architecture I had planned out looks a lot like this - client gets assigned a phone number from the pool for all its interactions with a certain business.

Good write up of a tricky problem, and glad to real-world validate the solution I was considering.

All the apps i've worked on lately in Rails use GoodJob, which is a Postgres NOTIFY/LISTEN based queue system.

Their offering was paid-for bundles of components and templates using tailwind, which they primarily drove traffic to via their documentation, which wasn't getting visited as much anymore because people just used AI.

Startups generally _don't_ end up with better outcomes. Large companies stay stable, startups are volatile and often end in failure.

Stability means removal of volatility, which means to stay stable they end up becoming more generalised, rather than the laser focus a small team like a startup can have. That laser focus can work out when applied to the right problem at the right time, but is very much not a guarantee.

I'm an software engineer with 17 years experience and I can't even get an interview at most places I put my resume in to.

Turso Database is a project to build the next evolution of SQLite in Rust, with a strong open contribution focus and features like native async support, vector search, and more

Some things you don't know people need until you're directly affected. For me, it was an injury related light sensitivity that made me realise dark mode isn't just a frivolous addition for looks

It shows the guy going to the restaurant, the same guy that eventually shows up at my door. It shows it on the way to a couple of deliveries and takes as long as extra deliveries should roughly take. It shows the immediate previous delivery when it's almost delivered, and the guy spends about as long as i'd expect at that place.

Not saying that it's not deceptive in some way, but it's more than just a surface-level difference.

When I choose priority delivery in Uber, I can see the driver go to the store, pick up my order and drive directly to my place. I also see the driver usually have 1-2 stops on the way if I don't select that. If there's enough gap between myself and the restaurant, priority is absolutely a time save.

If this is Uber then it's not legitimate.

I agree!

There's certainly a difference between making useful content for the love of it and making content because you think there's an opportunity to get something out of that (that could be money, but it could also just be appreciation or someone reading your work).

It's demoralising to not get any views on your hard work, and in this economic environment it sometimes feels more worth your time to do any other activity.

You may be the counter-proof to that and I enjoy your blog! But, also a lot of what makes your content useful is timing with depth and that's something that AI can't beat yet

AI scrapes niche blogs, Google deranks or spam drowns them out. It's really not a good time to be starting niche blogs.

Rocky's first release happened in 2021. It's presently on its 10th.

I'm going by the text on the page and didn't dig any deeper. On that page it says it's unreleased. Looking at the other comments it's not even the main site. No idea what this site linked is at all - maybe an older version where it was previously hosted, or some sort of malicious attack.

Any system that can respond to HTTP requests can be used as a web hosting system.

By this I mean its primary function is as a web host, or some kind of software that has a primary purpose of hosting websites. It's not Nginx, or Render or Vercel. An OS can host web software that can host websites, but not specialised for that purpose.

I would not be shocked that a website for general purpose software (like an OS) was hosted on a platform that removed the operations work from their plate, like Github pages or Vercel. They don't need it to be a demo of their work, and it's frankly not a good one if it were.

I'm not sure what the deal is here - from what i'm reading, it's an unreleased OS, not a web hosting system.

Again, not all replacements need to replace 100% or even 10% of plastic use to be able to have an a positive impact. There's space for a short-life plastic just like there's (currently) reasons for long-life plastics

Agree, but I don't see any mention of that in the article, so I don't have enough information to argue for that.

I'm sure we can agree though that having 17-day decomposing plastics that don't contaminate with heat and water is a good thing, so I hope it is that.

My point is it doesn't have to be a complete solution to replacing plastic to be able to have some benefits to replacing some plastics.

You can have local manufacturing processes so that it doesn't have to get stuck in a truck in Texas for a month.

And there'll still be uses for the long lived plastics. You don't have to use one plastic for everything - like we don't today.

Building a box that can last for centuries when you're only going to use it for 25 minutes and toss it is pretty wild if you think about it.

It doesn't _only_ solve long-term logistical problems. Plastics are used for things like takeout containers, drink cups and straws, amongst others - things that are only needed for a short time.

As far as I can tell, this is running from their servers - which means yes, you need to be able to trust the person who ultimately controls this at a bare minimum. Some trust him, some don't - some have good reasons, some don't.

I can evaluate this as it is, but if I was not trusting of a company, I can't then entrust my data to them, and so I can't evaluate a thing as any more than a toy.

This is pretty similar to the contemplating rewriting an app. There's usually a lot of knowledge and nuance built into legacy software that was hard-won, edge cases that are dealt with. Rewriting the app may be a good idea, but you have to ensure you're not losing what the legacy learned.

You can certainly make one-off apps to deal with things as they come up. But unless you are already an expert on what you need, you will still spend time building (or vibe coding), reacting to domain holes that you could also just spend $20/mo to ignore entirely.

The gap is smaller for sure, but it's not gone.

I don't think it's at all surprising that the injection of identity politics into the software industry has had a negative effect on quality.

That's a pretty broad claim. This conference could be in response to a perceived negative effect on quality, but claiming that as a fact seems hard to back up to me

It's been a good 16 years, though.

So the issues are:

* Your team were reluctant to use this AI tool, yet did so because you put effort into getting them to

* Anthropic was preventing you from using/signing up(?) and they were using this prevention to manage their load, which was apparently too high

* Your team, already not really wanting to use it, didn't use it because of a 1-time hurdle

* You are out money because of this

Is that the summary?

Pruning is exactly what you're looking for in a gardening SLM

There's other tells, like this other top HN post right now tries to work around: https://wbenny.github.io/2025/06/29/i-made-my-vm-think-it-ha...