I’ve been using it for a hobby project targeting a Linux server deployment. It’s good, and I don’t think the Apple ties are what’s holding it back now, I think it’s a relative lack of good frameworks and libraries. What’s there is often excellent quality, but it’s still a bit hit and miss even for common library use cases.
I often think about what my day-1 stack would be for a new company. Swift would be on the table as an option.
Indeed, it’s technically possible, but the level of discomfort, workarounds and paper cuts to get there makes it entirely a hobby thing.
If you really want to make “something” for the purpose of using Swift go ahead. If you want to make something, there are plenty of attentives to choose from.
No one at Apple uses Siri, that would be my guess and it provides no value to the company. Heck I'm surprised when anyone uses voice assistance in general.
Voice assistance was the hyped thing sometime before block chain and LLMs. Everyone needed one or was working on one. So Siri got stuck into the iPhone, like Apple Intelligence is now, because "the market" needed Apple to do so to keep stock prices going up. Then the hype died out Apple sort of needed to keep Siri around, because it is one of the interface to the HomePod and CarPlay. There's no reason do develop Siri beyond that, because very few people use it, not having a better version isn't going to sell less iPhone and they can't mine the data, because they sold everyone on Apple being privacy friendly.
There's absolutely no business case in having Siri do anything beyond "Call my wife", "Next song", "Play U2".
People don't use it because when it was first introduced, it responded "I found some web results for that" for the complex types of things people wanted to ask it.
The technology exists for much better responses to those types of requests today, but who knows if they can re-train their customers that it'll work for them now.
There are definitely a lot of complicated personal-assistant type tasks that they could do now that people would like.
Windows Vista happened and Microsoft still exists. And many other products from huge companies failed hard. I think these things are quite common for large companies. The fact that it's Apple just makes it more interesting.
Honestly, Windows Vista was a victim of the computing landscape of the time. Virtually none of what made Vista hated was changed at all to Windows 7, except by then everyone's PC had caught up to what the OS expected or they were going to stick with XP until the bitter end.
LLM just create a new layer of abstraction over the traditional "software" — and a very bad one. In a long run, this is more expensive than just doing "traditional" software development.
The first part of your sentence is entirely true, and the second part entirely wrong.
LLM-based programming values the prompting and architecture skills far above the code output itself, which is considered throwaway and can be replaced (and is replaced) often and quickly.
This method turns out to be much cheaper and faster than traditional development.
