Apple's week March 14: Absent Intelligence

Normally, I'd summarize some of the week's news and cover several of the smaller pieces, but there's nothing to discuss beyond a single event. Apple delayed the app intents-based feature for more contextual Siri and Apple Intelligence.

It wasn't really a shock to hear about the delay beyond Apple admitting it out loud. And as I write, no public statement has been shared from Apple about why they announced a feature or ran ads on it when it wasn't ready for prime time.

An internal meeting leaked where the Siri lead called the delay embarrassing, but that's about all we got. Based on that, it seems Apple did know the feature wasn't ready during WWDC 2024 where it was revealed or when the ad was run later that year.

I wrote over 5,000 words on the subject over the weekend then talked for 30 minutes on the AppleInsider Podcast about it. It's complicated because there are a lot of angles to observe. From the outside, it seems Apple is behind because all of its competitors are talking about society-ending AI and Apple is talking about Genmoji. The people that believe the AI lie look at Apple and say, “Is that it?"

Sure, Apple is behind the competition, but it isn't due to neglect, lack of data, or lack of prowess. It seems to mostly be about priorities. Apple began research into AI as early as 2016, and that's the research we know about. That's the same time OpenAI started research into generative technologies too.

However, Apple clearly thought the technology was going another way. It leaned into making its devices more efficient, taking better photos, and predicting what the user needs. The competition leaned into copyright theft to aid in developing lying machines.

From my point of view, Apple Intelligence is where it should be, if not a little early. I'd have been fine with them ignoring the stockholder pressure and releasing in 2025 instead of 2024. But here we are.

That brings the other side of this conversation – Apple's internal battle with AI. An excellent piece from John Gruber on Daring Fireball tears into the issues with Apple revealing something that didn't seemingly exist, then running an ad on it.

I feel like I could have written a very similar piece if I had approached from that angle. I didn’t, if only because my frustration is with AI as a whole and how it is marketed, not with Apple's approach. However, it is clear Apple made some uncomfortable and uncharacteristic decisions in releasing Apple Intelligence in 2024.

Let's hope this is just a blip in Apple history, no different than AirPower or Apple Maps. A black eye to the nerds who remember, but not a total internal alteration to the company's standards that leads to it becoming worse over time.

I truly believe AI is either at a ceiling today or will hit a ceiling soon as the technology currently works. There are always opportunities for breakthroughs, so I'm not trying to pretend I know something others don't. However, it just feels like nothing truly innovative has changed in the space in the last year.

There's a good chance that Apple will be able to catch up while everyone else struggles to find what's next for AI. Image Playground will inevitably allow users to make custom Memojis by describing them with text, Writing Tools will help users edit more complex syntax, and summaries will hallucinate less as it is honed into a better agent.

The true Apple Intelligence will be found in whatever Apple releases with personal context and app intents. As long as it isn't prone to hallucination or error, it'll make a material difference in people's everyday use cases.

For now, Apple Intelligence is making a difference for me, even if it's small, and that's enough. I never expected Apple to take over the world here, nor do I need any other company to do so. I really wish OpenAI would stop trying to suck up to the US government and stop pretending it's going to invent sentient life. It's an embarrassment to the technology.

Outside of the tech community, AI is just some kind of unknowable blob of annoyance. People don't know or care what it is as long as it's making things easier. So the first one to introduce AI that's actually useful, private, and secure will win. And in that race, Apple is in the lead.