Apple’s AI Isn’t the Problem. AI Is.

Apple has been taking a beating in the press for its messy AI rollout. The much-hyped AI-powered Siri got delayed indefinitely. The features that did ship, like text message summaries, are borderline useless. Critics are having a field day.

And look, some of that criticism is fair. Apple fumbled the launch. But the narrative that Apple is “failing AI” misses the bigger picture.

Let’s be real about why Apple and every other tech giant is rushing AI into their products. It’s not because users are begging for it. It’s not because there’s some clear problem AI solves that people are desperate for. It’s because Wall Street wants a “super cycle” — a new feature so compelling that millions rush to upgrade their devices.

Investors have been salivating over this idea for years. Apple, desperate to please shareholders, tripped over its own feet. And now the company is quietly walking back its promises, saying the delayed features will come “in the coming year.”

That’s fine. What bothers me is the framing that Apple has somehow failed AI, rather than the other way around.

There’s this weird dogma taking hold in tech circles: AI can never fail, it can only be failed. Failed by users who don’t understand it. Failed by companies that aren’t aggressive enough. Failed by regulators who hold it back. The technology itself is always blameless.

Kevin Roose from the New York Times made this exact argument on his podcast recently. He said Apple needs to get more comfortable with error, with mistakes, with things that are “a little rough around the edges.” That Apple should just push AI into products and accept that it won’t be polished.

I respectfully disagree. Hard.

Apple built a $3 trillion empire on the opposite philosophy. The company is famously obsessive about detail. Its “walled garden” approach to iOS drives developers crazy, sure, but it’s also why a billion people trust Apple with their faces, their bank accounts, their real-time locations. You buy an Apple product and it just works. Your boomer parents can figure out FaceTime without a manual.

Now we’re supposed to accept that AI features should ship half-baked because the technology is too important to wait? That users should learn how to tiptoe around the limitations of large language models that might or might not give accurate answers?

To what end?

As Roose’s co-host Casey Newton pointed out in that same episode, Google and Amazon haven’t figured out some incredible use case that’s making people rush to buy new Pixels or Echo speakers either. AI is still much more of a science and research story than a product story.

Large language models are fascinating. They’re an academic achievement with real potential. ChatGPT and Claude have found commercial success. But a bot that’s 80% accurate isn’t a product. It’s a prototype.

Apple’s sin isn’t failing AI. It’s treating AI like a feature that customers actually want, when the evidence suggests most people don’t care. The company stumbled because it tried to satisfy investors instead of users. That’s a strategic error, not a technological failure.

The tech industry needs to stop pretending that AI’s shortcomings are everyone else’s fault. The emperor isn’t naked. He’s just wearing a prototype that doesn’t fit yet.