top of page

WWDC 2024: Apple Joined the AI Race



As an Apple fanboy, every year I sit in front of the TV during Apple announcements like WWDC and the various product launches, the way a sports fan sits in front of the TV for the Super Bowl or the Premier League championships. This is my time.

Apple has a way of introducing and announcing products and features like no one else. They can make something so basic seem like a big deal, or announce a feature that Android had 10 years ago, but make you feel like it's a great innovation. Regardless, I love these events. Sometimes they're a hit, sometimes they're a miss.

The AI Anticipation

Over the last year, as I've been diving deeper into AI - theoretically, practically, GenAI, and more - I've been eagerly waiting to see what Apple would do. How would they compete with Google? How would they integrate the amazing advancements we've all been enjoying from companies like OpenAI, Anthropic, Stability, Mistral, and more?

During last year's WWDC, when the recent GenAI craze was already in high gear, Apple announced the Vision Pro - an impressive device that's way out of most of our justifiable price ranges, especially for a gen1 product. Yet, they didn't mention "AI" whatsoever. It was like an alternate reality; we were all waiting for it, but it didn't happen.

As the year progressed and companies like Google and OpenAI seemed to be head-to-head in this AI arms race, Apple remained silent. Don't get me wrong, Apple has been silently incorporating neural networks and GenAI features for years - small things like improvements in Photos, predictive text, smart grocery lists, to name a few. But they remained silent about any major AI initiatives.

WWDC 2024: Rushing through the details

So WWDC 2024 rolls around, and we all know it's coming... Wow, a 2-hour event! In the first hour, they're speeding through all the usual long detailed updates to Vision OS, iOS, watchOS, iPadOS, tvOS, and macOS changes, but this time they're powering through as fast as possible. I couldn't help but feel a mix of excitement and mild underwhelm. Don't get me wrong, there were a few features that really got me excited, some things we've been waiting for a long time. But it just felt to me like they were rushing to get to something bigger.

Some of the features I'm particularly excited about:

  • Enhanced privacy controls on iPhone

  • New vitals tracking on Apple Watch

  • Ability to take rest days without losing streaks (can I get this retroactively, please?)

  • Controlling other people's phones via FaceTime (I cannot begin to explain how much easier it will be to help my dad with his phone without losing my mind)

  • The new notes, math notes, and handwriting features with Apple Pencil

  • Complete control of your phone from your computer

  • Scheduling messages for later

The new meaning of AI reveal

And then Apple did it. They rebranded AI to "Apple Intelligence" (btw Alibaba did this too not too long ago). But more than just a rebranding, their approach to AI hammers down on one of my biggest concerns about the current world of AI: privacy.

One of my biggest worries when using ChatGPT, Claude, Perplexity, or even one of the many awesome open-source models is privacy. What's happening with my data? Should I really upload this medical document? Will it be used for something else?

This is where Apple's walled garden approach and commitment to privacy are really starting to pay off. Apple's silicon is proving its worth, and their dedication to leading a privacy standard is what truly matters in this new AI landscape.

Apple's Approach: On-Device AI and Privacy

What really stood out to me during the WWDC presentation was Apple's emphasis on how their AI features will be implemented. They made it crystal clear that privacy is at the forefront of their AI strategy, and here's how they're making it happen:

  1. On-Device Processing: Apple explained that the vast majority of AI features will happen right on your device. This is huge. It means your personal data, queries, and interactions with AI stay on your iPhone, iPad, or Mac. No need to send sensitive information to external servers.

  2. Apple's AI Data Centers: Now, here's the kicker - if a task is too complex for your device to handle locally, Apple doesn't default to sending your data to a third-party cloud. Instead, they've built their own AI data centers. They're calling it "Private Cloud Compute" which is apparently powered also by their own apple silicon (M series chips). So even when processing needs to happen off-device, it stays within the Apple ecosystem.

This two-pronged approach is brilliant for several reasons:

  • Privacy: Your data isn't being shared with or processed by third-parties.

  • Speed: On-device processing can be incredibly fast, with no need to wait for server responses.

  • Offline Capabilities: Many AI features will work even when you're not connected to the internet.

  • Personalization: By keeping your data local or within Apple's ecosystem, the AI can be more personalized without compromising privacy.

The role of Apple Silicon in this cannot be overstated. It's like Apple has been laying the groundwork for this moment for years, and now we're seeing why.

This approach stands in stark contrast to many other AI implementations we've seen. While companies like Google and OpenAI have made incredible strides in AI capabilities, they rely on cloud processing. Apple's strategy of prioritizing on-device processing and using their own data centers when necessary could be a game-changer in terms of privacy and user trust.

The ChatGPT Curveball (We All Knew Was Coming)

After 40 minutes of hammering home how everything happens on-device or on their secure servers, they casually dropped a bomb: Siri can also integrate directly with ChatGPT. Wait, what?

Although this partnership was already leaked over the past few weeks, we didn't know what it really meant in practice, and overall the announcement itself felt like an afterthought. We just learned about all the AI integrated throughout the Apple ecosystem, but on top of that - we can now also have the power of ChatGPT right there in Siri. To be honest, it's a bit confusing at first, mixed messaging, I'm sure it will all be explained better, but it raises some big questions about privacy.

Apple spent the better part of an hour telling us how they're keeping our data safe within their ecosystem. So how does ChatGPT fit into this privacy-focused approach? Will our queries be anonymized before they reach OpenAI's servers? Will we have control over what information is shared? Or will Apple somehow run ChatGPT models on their own hardware?

If Apple can figure out a way to harness ChatGPT's capabilities while maintaining their strict privacy standards, it could be a game-changer. But it's a delicate balance, and I'm really curious to see how they pull it off.

Also, it's not fully clear to me yet where the abilities of Siri lack and will need ChatGPT to pick up the slack.

This unexpected twist adds another layer of complexity to Apple's AI strategy. It shows they're not afraid to collaborate with other AI leaders, but it also puts their privacy promises to the test. I'll be keeping a close eye on how this integration works in practice. Only time will tell.

The AI Features I'm Pumped About

The Apple Intelligence features I'm most excited about:

  • A smarter, more context-aware Siri

  • Rewrite capabilities for text

  • Summaries of everything, including emails, notifications and more

I'm excited because in theory, Apple's own AI will finally fix Siri, give it much more intelligence, agency, and the ability to do things for us. Giving Siri context is huge. Are these the Agents we we’ve been looking for?




Apple is also adding their own image generation and writing capabilities on-device, built natively into all the apps we love. And then you can use ChatGPT to do even more. I guess it's like Siri will be powered by their own LLM or the SLMs they released recently, but then the (hopefully) secure ChatGPT integration makes it F - -ing awesome.

Conclusion: Apple's AI Gambit

As the dust settles on WWDC 2024, I can't help but feel a mix of excitement and anticipation. Apple's entry into the AI race isn't just another tech giant jumping on the bandwagon; it's a calculated move that could reshape the entire landscape of artificial intelligence.

What stands out to me is how quintessentially "Apple" this approach is. They didn't rush to be first, but they've entered the game with a solution that addresses one of the biggest concerns in AI today: privacy. By leveraging their ecosystem, from on-device Apple Silicon chip to their new Private Cloud Compute, they're offering a unique proposition - advanced AI capabilities without compromising personal data.

Of course, questions remain. How will Apple's AI capabilities stack up against the likes of GPT-4o or Claude? Will the on-device processing be powerful enough for complex tasks? And how will developers be able to leverage these new tools?

The most intriguing aspect is the potential ripple effect across the industry. If Apple can demonstrate that powerful, user-friendly AI can be implemented with a privacy-first approach, it could push other companies to prioritize privacy in their AI developments. This could lead to a new era of AI where advanced capabilities don't come at the cost of personal data.

I can't wait to get my hands on these new features (dev beta, here I come, weekly panic attacks and all). More importantly, I'm eager to see how this moves the needle in the broader AI landscape.

The AI race just got a lot more interesting, and I, for one, am here for it. Bring on the future of Apple Intelligence!


Comments


bottom of page