willtemperley

Thanks that’s an interesting anecdote to me because I’m also developing SwiftUI apps. Claude will create something that technically works but looks terrible.

I don’t have the luxury of a previously well designed app to train Claude on, but I’m finding just building the UI the old fashioned way then asking Claude to fix the bugs works to an extent.

I’m hoping that creating a comprehensive design system with good semantic naming will help.

I think this is true to an extent, but it really depends on how much information the LLM has on a topic.

They’re amazing at maths because maths has been open source since day 0. They’ve great at algorithms for similar reasons. C bindings are a doddle because there’s so much prior art available. They’re terrible at using cutting edge features in languages because there’s not much data yet.

I think it’s fairly easy to predict what they’re good at on this basis. I’m not sure why they suck at UI design though.

They wouldn't, but build scripts are a particularly effective attack vector.

Surely non-sandboxed build scripts are just a terrible idea.

Both Cargo and npm should look at what Swift Package Manager (SPM) is doing.

It’s not perfect but there’s a noticeable absence of supply chain attacks involving SPM, probably partly because it doesn’t use a mutable registry, but I suspect attacks are just more difficult. On the rare occasion a build script is involved it’s run in a sandboxed plugin.

Agree on the styling, the diagrams are very clear and match the text perfectly. I like the trailing dots though.

Yes my grandfather had one. It worked well on flat terrain. I think it used a motorcycle battery so batteries were easy to find and charge.

I suspect you're not a pretty girl that gets stared at. Neither am I, but if I were I'd rather the stares weren't recordable.

> "Claude did it" cannot ever be a valid excuse.

I think this an incredibly important point. As soon as we devolve reposnsiblity to computer progams, society cedes control to the algorithm vendors.

Jeff Geerling was called out today for his clickbaity video entitled "Intel just matched Apple Silicon. Seriously.". Jeff's personal response was "I had three YouTube titles, this one got about 2x the watch time as the other two, so I let YT pick it."

Another commenter replied: "You are trying to launder responsibility for a clickbait title to a computer program." I think this exactly nailed the problem.

https://news.ycombinator.com/item?id=49223079

Yep, “instinct“ and “smoking gun” are LLM favourites.

I find it really annoying when the LLM says “good instinct” as if I’m an animal barely able to think.

I'm against LLM usage in conversation, e.g. when discussing issues in open-source. It's like half-hearing a conversation, I'm missing the context that the other person has with their robot. Second, the person has usually at best half-understood the issue.

Provided the code does the job, it seems immaterial whether it was hand-written or LLM generated, provided it does the job and isn't straight-up plagiarism. The latter is admittedly a hard thing to judge.

The Cornetto is truly important in British culture. For example, the Cornetto trilogy, which i'm surprised wasn't mentioned:

https://en.wikipedia.org/wiki/Three_Flavours_Cornetto

GNOME 2 was lovely.

Then GNOME 3 came and I switched to Mac.

With all the myths and hyperbolae circulating regarding AI, I'd love to know what it's like at large software companies adjusting to this brave new world.

It's easy for a small team to adjust workflows and roles, but I just imagine the office politics must be a waking nightmare in big organisations right now.

I tried that but Claude thought I was already outstanding in my field.

I would only take this article seriously if you have to support old iOS versions.

NavigationStack has been around since iOS 16, @Observable since iOS 17. and I'm not sure if the author is aware that GeometryReader hasn't been necessary since iOS 18 added onGeometryChange modifiers. Nearly 100% of users are on iOS 16+.

I'm doing some fairly complex work with SwiftUI and I find most components work identically between iPad and macOS. I can work for a week without needing to test on an iPad.

There are problems with SwiftUI for sure: I'm finding macOS performance on an M2 Max is far worse than an iPad M1, especially with animations. Toolbars are very inconsistent between platforms and the compiler timeouts are a real pain (hopefully Swift 6.4 will fix that).

> The latest is NavigationStack, which is still pathetic. The only way to navigate more than a level deep is to create an array of one datatype to serve as NavigationStack's "path." But this is designed to be an array of a single datatype.

You can use a tagged union for the "one datatype", thereby allowing your views to accept any datatype you provide in the enum. Or use NavigationPath.

No.

What isn't reasonable about this approach? I think most people have got GeoAI backwards, creating derived imagery via AI is a terrible idea. Building better tools than we have with AI, using existing geoscience principles is absolutely the right approach.

> how is anyone going to determine what is real or fake anymore

Through using AI to build tools that generate results with a verified chain of provenance.

This is absolutely the headache I’m having today, an enthusiastic contributor pasting AI generated wall-of-text comments to justify their 500 line PR.

How can we have a conversation and reach a shared understanding if one of the participants is avoiding attempting to understand the problem?

I have no problem with AI generated code, provided it’s reviewed, but I might as well be talking to a brick wall if the person hasn’t understood.

My jaw drops every day.

Small change in comparison, but today Fable created and benchmarked an RTree which was 100x faster to populate and 5x faster to query, compared to a previous attempt with Opus 4.6 a couple of months ago. That took about two coffees.

I think it's important to remember that as impressive Opus & co are, they're standing on the shoulders of giants, i.e. the engineers, academics and companies who have cooperated to design the incredible programming languages and machines we have today.

Yes it does sound like a value judgement made by someone who "fits in", probably after drinking several pints down the pub.

However I think the poem may be deeply ironic, given Robert W Service was probably a member of the "Lost Legion" himself:

https://www.poetryfoundation.org/poets/robert-w-service

> I don’t want to be sold the same thing over and once again.

I don't want to be sold the same music, algorithmically recreated again and again to maximise profit while circumventing copyright law.

A lot of people like analogue music. I certainly do, and I wonder if Korg are anticipating a reaction to the slop.

I wonder if this is this just the law of diminishing returns at play?

My thinking being, it's been a few months since I thought the code generation machine was the problem, rather than my interactions with the machine. A month is a long time in AI.

What I mean is, these things are about as smart as they need to be already for the average SWE. I don't think this is true for those solving the really big questions like curing cancer.

Also on a macbook, it took me a while to realise pinch to zoom doesn't do anything except zoom the page, pan is click and drag, and scroll is zoom. Cool project except for the controls, surely scroll should be to scroll up and down?

Looks like Apple are making an effort to get agents to create decent code with Xcode skills, which someone has extracted here [1] but there is near zero mention of accessibility. Worth a read though I think, I learned a few things.

[1] https://github.com/superagents-lab/xcode27-skills/

Vim is a bad example to use here. I certainly have no hacker-vibe pretensions, I use Xcode almost exclusively, but with the Vim key bindings which have been integrated very well. The productivity gains with Vim commands are impressive.

Yes. I'm not sure whether Amodei's motivation for attempting to ban open-weight models was for safety or to strengthen Anthropic's position though.

I do hope Anthropic are better with IP, and I think they may be. Given Dario Amodei hasn't been sued by OpenAI while building Anthropic this seems likely.

I think Amodei may actually be quite a good human, despite my trust in big tech being at an all-time low.

This is a really bad look for a company that has vast quantities of our IP stored on its servers.

Can you provide a source? Otherwise your comment is useless.