cageface

Verified this locally myself. Thanks for the concrete test. I guess it's time to give Claude another try.

Codex does better if you ask it to take screenshots and critique its own UI work and iterate. It rarely one-shots something I like but it can get there in steps.

In the US. In many other countries it's borderline useless.

LLMs tend to do better with component libraries like shadcn or daisy than writing raw CSS in my experience.

The pro plan is useless. You need at least the 5x max plan to get any real work done.

That said I find the GPT plans much better value.

If you think Android tooling is subpar wait until you try iOS.

I've had exactly the opposite experience. Getting great results using GPT for hours every day since 5.3. You need to put the effort level on at least high though.

Every time I hand off a task to Opus to see if it's gotten better I'm disappointed. At least 4.7 seems to have realized I have skill files again though.

On at least high effort level I find GPT 5.4 easily beats Opus 4.6 in code generation and debugging issues.

Unless you need to work with a repo that uses submodules or lfs.

That hasn’t been my experience. I agree Opus has the edge but it’s not by that much and I still sometimes get better results from Gemini, especially when debugging issues.

Claude Code is much better than Gemini CLI though.

I really don't know how effective LLMs are at that but also that puts you in an extremely narrow niche of development, so you should keep that in mind when making much more general claims about how useful they are.

Why is that the real metric? If you can turn a 1x dev into a 2x dev that's a huge deal, especially if you can also turn the original 2x dev into a 4x dev.

And far from "churning out code" my work is better with LLMs. Better tested, better documented, and better organized because now I can do refactors that just would have taken too much time before. And more performant too because I can explore more optimization paths than I had time to before.

Refusing to use LLMs now is like refusing to use compilers 20 years ago. It might be justified in some specific cases but it's a bad default stance.

C, Swift, Typescript, audio dsp, robotics etc.

People always want to claim what they’re doing is so complex and esoteric that AI can’t touch it. This is dangerous hubris.

The denial on this topic is genuinely surreal. I've knocked out entire features in a single prompt that took me days in the past.

I guess I should be happy that so many of my colleagues are willing to remove themselves from the competitive job pool with these kinds of attitudes.

This kind of take I find genuinely baffling. I can't see how anybody working with current frontier models isn't finding them a massive performance boost. No they can't replace a competent developer yet, but they can easily at least double your productivity.

Careful code review and a good pull request flow are important, just as they were before LLMs.

This is a large exaggeration. There are still a lot of people working on Flutter at Google and large companies continue to adopt it.

Google's wildly popular NotebookLM is a recently released Flutter app, for example.

I’ve had very good experiences with Flutter on iOS and macOS. It’s actually a lot easier to get good performance in Flutter than SwiftUI.

No cross platform stack can do Liquid Glass yet. You have to wonder if that was one of design goals.

The last time I looked at it was far less mature on non-Android platforms than Flutter. Has that changed?

You could give my app a try:

https://www.plastaq.com/minimoon

SwiftUI on macOS 26 still has issues but it’s finally starting to evolve into something usable. In particular it seems like the long standing performance problems are being addressed.

As another example, the MacApps Reddit has been flooded with new apps recently.

I'd add postgres to that list.

Their Cocoa based frameworks were good for their time, I'll grant you that. Swift & SwiftUI, at least in their current state, are a mess with some pretty fundamental issues.

I agree but I also think any meaningful regulation is off the table for the next few years in the USA at least.

This matches my experience. Codex is just more capable. But Sonnet is also quite good and much faster.

So lately I'll start with Sonnet for everything but the most complex tasks and then switch to Codex when needed.

For writing code at least this has been exactly my experience. GPT5 is the best but slow. Sonnet 4.5 is a few notches below but significantly faster and good enough for a lot of things. I have yet to get a single useful result from Gemini.

This has not been my experience with Flutter at all. It's made building a non trivial cross platform app so much easier and for the few things I've needed to drop down to native it's been very easy.

He very often brings other knowledge to bear in conversations that clearly wasn't just prep work for the specific topic. He's not strong on science but he has a firm grounding in literature, philosophy and history that isn't common these days.

It won't be easy to find somebody with his breadth of knowledge to replace him.

I finally gave up and pulled all my iOS apps from the store after receiving yet another app update rejection from a reviewer that clearly hadn't spent 5 minutes trying to understand what the app is for. I'm tired of begging for having access to my users restored every time something like this happens and I'm through feeling that powerless.

From now on I'm an Android user and only building PWAs.