Apple's right to take it slow with generative AI

With Apple’s annual WWDC event just around the corner, rumors surrounding the new versions of iOS, iPadOS, and macOS have flooded the internet. While there are a lot of different details to pick apart, one proclamation stuck out to me — that Apple choosing to launch its upcoming AI features in beta is some kind of proof that Apple’s unprepared, unready, or caught off guard by the AI race. Let’s get into it...

I’ve already covered the ramifications of OpenAI’s dystopian future and Google’s internet-breaking AI Overview feature extensively on ZLR. Despite the flashy presentations and promises of better tomorrows, all it takes is a bit of attention to realize that even the two top generative AI solutions on the market are flawed beneath their fancy facades.

Parade of fools

For instance, just earlier this month, the CEO of OpenAI, Sam Altman, called its now-defunct premium ChatGPT 4 platform – the same version that customers paid for the privilege to use – “mildly embarrassing at best” and “the dumbest model that any of you will ever have to use again by a lot.” Those are sharp words from a man who once willingly charged customers a non-small monthly fee for access. OpenAI has since launched a new, much smarter, and potentially dystopian model, dubbed ChatGPT-4o, which you can actually try out for free before getting an alert to pay up.

Google was also quick to get in on the action. Within hours of launching its controversial AI-powered search solution, aptly named AI Overview, users prompted it to serve up obscure Reddit comments to queries and deliver them as fact. Some of the best examples involved putting glue on pizza and eating a hearty helping of rocks each day, among other recommendations.

In addition to that, an investigation late last year uncovered that “ChatGPT technology makes things up about 3 percent of the time” when summarizing facts – like the time it completely fabricated court casesthat were referenced by an actual lawyer in a real trial. Gemini scored even worse as “Google system’s rate was 27 percent.”

In other words, AI solutions aren’t always as reliable as they’re marketed to be. And unless the responses they give are obviously bizarre (as in AI Overview’s case), regular users won’t always know when the information presented is accurate or not.

That brings us to Apple.

Apple’s right to put AI in beta

Apple has undoubtedly paid attention to the rapid pace of AI development and the missteps companies like OpenAI and Google have taken in an effort to beat each other to the punch. As a high-profile brand with a valued reputation, Apple is unlikely to move too far, too fast in a way that might damage its luxury image.

On top of that, Apple’s never been quick to jump the gun on any new product or service. They famously wait for certain technologies to mature before releasing their own versions.

For this reason, Apple isn’t unprepared or unnecessarily cautious for launching its upcoming AI features in beta. Quite the contrary, this is simply the responsible thing for a multi-trillion-dollar tech giant with billions of global customer to do. Apple wants to make sure they get their AI execution correct before fully unleashing it on customers, and that is a good thing.

A cautionary tale

That said, Apple is still very obviously behind the ball when it comes to generative AI for another reason. The timing of it all suggests that they simply didn’t see it coming down the pike. As Apple focused on AR/VR with its next important product vector (which we now know to be the Vision Pro), the tech market was rapidly shifting toward AI as the next big thing, and it’s only picking up steam.

However, while the fast evolution of AI has introduced some mind-blowing capabilities, there have also been plenty of huge missteps from key players that prove why Apple is correct to proceed with caution. Every tech giant should be more mindful of the power that’s being handed to users through AI, both the good and the bad.

As a fan of Apple’s products, it’s good to see the Cupertino giant taking their time instead of trying to cram AI into every facet of their operating systems before any of us really know what AI will do to society in the long run.