[pllbnk]

I agree. As usual these days, the culprit for lack of innovation is AI (mostly LLMs). Apple did try to shove it to the users unsuccessfully and they had the right amount of sanity to backtrack once they realized how the users' expectations are misaligned with what LLMs have to offer, which is mostly lack of reliability; average user thinks that something walks like a duck, quacks like a duck must be a duck, but LLMs, while pretending to be human-like are not like that. On the other hand, Microsoft is still forcing it upon their users with Copilot but users are resisting. We will see which tactic will win in the long-run but I would bet on Apple.

[lubujackson]

If history is any indication, Microsoft will slowly improve until things become sort of decent, then Apple will release a neutered but polished version that is lauded as "Apple Invents AI" and immediately dominate the market.

But their version will lack one critical feature that everyone expects, such as a power button, or copy/paste, or pinch to zoom. For AI, maybe something like "remember context".

Then 3 years later they will invent "remember context," and everyone will lose their shit again.

[NewJazz]

The article addresses Apple's historical late mover advantage strategy, and suggests that Tim Cook doesn't have the same vision that Jobs had to pull those moves off.

[IOT_Apprentice]

Cook isn’t a visionary, he is an ops & logistics guy. He has a major problem with failing at AI and lying last year about “Apple Intelligence”.

He needs some fresh blood in software as well.

He should just buy something like Claude and lock them in with stock and very high salaries.

He should also stop scrimping on integrated memory on the M series. Overpriced memory is a huge issue. And do something about the GPU as well.

Perhaps an AI N series with 128gig of ram, & dramatically more powerful GPU with higher core/thread count.

[danielscrubs]

Buying competency and hoping for the best sounds exactly what an economist would do.

They will probably leave as soon as their manager becomes just the run of the mill babysitter manager with charisma.

[Nevermark]

So true. That was Meta's all too generic strategy for the Metaverse, throw tons of money and people at the problem. And it worked as poorly as one would expect.

Now Meta is doing it again for AI.

Apple won't throw money down a hole like that, but they definitely need a more original and opinionated strategy than just an increased talent count.

The lack of love they gave Siri and AI in general, since they bought the original Siri tech in 2010, just keeps looking like a worse and worse oversight.

[delfinom]

Ah yes, Jobs famously could solve the LLMs being shit problem overnight.

By convincing users they want hallucinating, gas lighting imps in their pockets 24/7 to reinforce their thoughts.

And he would have been all over it given his own alt-medicine beliefs.

[johnisgood]

He is dead, we cannot be certain.

[tim333]

>...backtrack once they realized how the users' expectations are misaligned with what LLMs...

Not my experience as an Apple user. The issue was not that I expected something better than an LLM. It was that I'm used to ChatGPT and the Apple AI was rubbish in comparison.

And if did useless stuff like mangled summaries of messages. If I could set if to know what files I have and web pages I'd used and could ask "which file do I have the holiday records in?" etc. then that would be useful.