Skip to main content

The internet is a sticky mess, and Apple just gave us an AI mop

A person demonstrating the new Siri revamped with Apple Intelligence at the Worldwide Developers Conference (WWDC) 2024.
Apple

If less is more, then yesterday’s AI announcement from Apple was both modest and well, by implication, huge.

Sure, you can use AI to create your own emojis, find the right images for a presentation, or transcribe phone calls, but the real theme of Apple’s long-awaited AI fiesta was the ability to remove content from your life. And I’m not talking about crushing centuries of human ingenuity into an iPad. Apple Intelligence, as Cupertino’s AI variant is called, feels more like a Marie Kondo for your iPhone than a Picasso painting — and it comes at a time that the internet’s clutter problem has never been worse.

Recommended Videos

The poison and the remedy

Unlike the chabot interface you probably know from ChatGPT and Google Gemini, Apple Intelligence is more of a series of features baked into Apple apps.

It will figure out which emails in your inbox need attention, summarize messy reply threads, and even help you be more concise as you tap out a reply. The intrusive notifications that swarm your iPhone’s lock screen can likewise be prioritized, summarized, and tailored by AI to be less disruptive.

It’s not just about combating the outside forces vying for our time. Even our own mountains of content can be popped in the AI sauce pan and boiled down to a rich nostalgia gravy: Memory Mode helps turn your mountains of photos and clips into a cohesive video with a prompt like “Leo learning to fish and making a big catch, set to a fishing tune.”

Using Apple Intelligence to create a Memory movie on iOS 18.
Apple

If it sounds like I’m dismissive, let me stop you right there: I want all of these features. From coupon codes to notifications of subscription increases, my inbox is as messy as anyone’s. And that’s just the beginning of the digital inundation most of us face daily. Throw on group text threads, reminders, plus all the bleeps and bloops from home security cameras, robot vacuums, and smartwatches, and it’s little wonder we need AI to sort the 13th low battery notification on the back door lock (yawn) from the alert that tomorrow’s flight time just changed (oh, crap).

We’ve applauded Apple for giving us AI tools we genuinely need. But it’s hard to miss the irony that AI is both the levee that protects us from the rising river, and the upstream sewage plant disgorging the very effluent lapping at its shores.

Just look at Google. Google Search has become so fundamentally plagued by scammers pumping it with low-quality AI content that it introduced AI Overviews several weeks ago to summarize the garbagescape for you. “Let Google do the Googling for you” is a real slogan, not a new spin on the old Xzibit meme. We’re so awash in content that we need content to tell us which content to consume.

The trash clogging Google’s arteries is created by voracious businesses trying to exploit AI. But you and I aren’t immune. As easy AI tools proliferate, AI will creep into our personal communications, too. Apple showed off AI-powered writing tools generating a poem for your housewarming invite.

A screenshot from Applke WWDC shows AI writing tools being used to write a poem.
Apple

Cute. But it’s hard not to imagine that the AI-generated poem is precisely the kind of crap someone else will use AI to filter out when they just need to know what time they’re supposed to show up with the pasta salad.

Which will we end up using AI for? It’s looking a lot like both. And if history is any indication, that’s not exactly unprecedented, or anything to fear.

It’s up to us

Like any technology, AI will be what we consumers make of it. Remember the halcyon days of GeoCities and Angelfire, when suddenly anyone could author a website? As nostalgic as I am for those animated “under construction” signs now, they were a mess. Our mess, created with whatever primitive tools we could get ahold of.

It took more, better internet tools — like early Google — before the internet could become a less chaotic place with real utility. Which is why Big Tech doesn’t get off the hook for the garbage monsoon generated with early AI: The way new technologies are handed to us matters. As different companies rush to AI-ify everything they touch, we’ve seen plenty of cringe-inducing junk by founders who clearly never made it past the look-what-we-can-do phase. And we’ve also seen some truly exciting game changers.

Its approach isn’t perfect, but Apple has been the best so far at crafting AI tools with clear intention. Hopefully, other companies take notice, and help usher this technology past its Pets.com era of hype and into something we can really use. Because regardless of how carefully some of us choose to deploy its wonders, we’re already behind a slime trail of greedy companies trying to wring every dollar they can from it. And we’re going to need a big mop.

Nick Mokey
As Digital Trends’ Editor in Chief, Nick Mokey oversees an editorial team covering every gadget under the sun, along with…
I finally tried Apple Intelligence in macOS Sequoia to see if it lived up to the hype
The redeisgned Siri user interface in macOS Sequoia.

For the last few years, Apple’s macOS releases have been interesting, if not particularly exciting. But that’s all set to change this year with the launch of macOS Sequoia, and it’s all thanks to one feature: Apple Intelligence.

Apple’s artificial intelligence (AI) platform has the potential to completely change how you use your Mac on a daily basis. From generating images, rewriting emails, and summarizing your audio recordings to revamping Siri into a much more capable virtual assistant, Apple Intelligence could be the most significant new macOS feature in years.

Read more
We just learned something surprising about how Apple Intelligence was trained
Apple Intelligence update on iPhone 15 Pro Max.

A new research paper from Apple reveals that the company relied on Google's Tensor Processing Units (TPUs), rather than Nvidia's more widely deployed GPUs, in training two crucial systems within its upcoming Apple Intelligence service. The paper notes that Apple used 2,048 Google TPUv5p chips to train its AI models and 8,192 TPUv4 processors for its server AI models.

Nvidia's chips are highly sought for good reason, having earned their reputation for performance and compute efficiency. Their products and systems are typically sold as standalone offerings, enabling customers to construct and operate them as the best see fit.

Read more
Apple is tackling one of the most frustrating aspects with AI today
Apple Intelligence on AI

As companies like Google, Anthropic, and OpenAI update and upgrade their AI models, the way that those LLMs interact with users is sure to change as well. However, getting used to the new system can become a hassle for users who then have to adjust how they pose their queries in order to get the results they've come to expect. An Apple research team has developed a new method to streamline that upgrade transition while reducing inconsistencies between the two versions by as much as 40%.

As part of their study, "MUSCLE: A Model Update Strategy for Compatible LLM Evolution," published July 15, the researchers argue that when upgrading their models, developers tend to focus more on upping the overall performance, rather than making sure that the transition between models is seamless for the user. That includes making sure that negative flips, wherein the new model predicts the incorrect output for a test sample that was correctly predicted by the older model, are kept to a minimum.

Read more