AI to power the corporate Windows 11 refresh? Nobody's buying that
Microsoft should look to Apple for lessons in flogging dead horses
Opinion In the early 2010s, Intel's PR did the tech press rounds with a hot story. We're so far ahead in chip fab, they said, that nobody will ever catch up. The hacks concluded two things from this: Intel was losing ground fast enough to be scared, and it was right on both counts. How did that pan out again?
So it is with the news that PC manufacturers are hyping up the long-delayed next big corporate hardware refresh. Michael Dell and pals rallied to the fact to advance three big reasons for this. The first? The last big wave of desktop replenishment was more than four years ago, long enough for those fleets to be clapped out and useless. Secondly, there's AI, which needs shiny newness, and you can't say no to that.
The third big reason is that Microsoft is winding down support for Windows 10, and Windows 11 is restricted so as not to run on old kit. Unlike the first two reasons, this one may be nearer the mark. Also unlike the first two, it is entirely negative for IT departments, like a pet shop saying they'll shoot your dog in the new year, so you'd better buy a puppy for Christmas. It's not clear this will actually work on this occasion.
The first two reasons are palpably false. PCs stopped getting old when hard disks went away. By then, CPU speed and memory size had ceased to matter on the company desktop – and beyond. You can buy a retired corporate PC from the last refresh for a two-digit sum on eBay, spruce it up a bit with a last-but-one-gen GPU and have yourself a decent gaming rig. Excel was never a problem.
If you were stuck at work in 2015 with hardware and software from 2005 – Windows XP, yes – you had a genuine grievance. If you're using a 2015 PC with Windows 10 in 2025, you may not even notice. As the Arch-Mage of Microsoft scrying Mary-Jo Foley says: "No clear reason/features pushing upgrades."
As for AI becoming the magic new feature to change that, Microsoft is wishing extra hard that it is so. It is not so. The corporate world regards the sort of AI that Microsoft and others are flogging much as a toddler eyes broccoli. It may turn up on the plate, but ain't nobody ordered the stuff.
There is a way forward, from Apple's parallel consumer universe. While Cupertino has the strong advantage of directly controlling both OS and hardware, it also has a lot more competition, at least on the surface.
Apple Silicon might seem to be another advantage, but it has yet to reach the handset, and it is here that things get interesting, because Apple has met the same practical hardware and OS feature stagnation that stunts the corporate refresh cycle.
The flagship of four years ago, the iPhone 12 Pro Max, still looks gorgeous, and more to the point it runs iOS 18, released in late 2024, perfectly well. Like the iPhone 16, it has no noticeable lag on its glorious screen. It is highly unlikely that an unschooled user could say with any degree of certainty which phone was four years old, and which was released yesterday. Yet the new series will sell, and likely sell very well, mostly to people already invested in the platform on their own dime. It's nothing to do with the promised but not in production AI either.
- China's quantum* crypto tech may be unhackable, but it's hardly a secret
- Upgrading Linux with Rust looks like a new challenge. It's one of our oldest
- Microsoft decides it's a good time for bad UI to die
- Sorry, Moxie. Blaming Agile for software stagnation puts the wrong villain in the wrong play
The constant incremental difference that keeps them coming is the camera system. People may not be taking better pictures and videos than they did with previous generations of iPhone, or against the opposition's flagships, but the new features and capabilities are instantly understandable and the results immediately appreciated.
Pictures matter to people more than most things a smartphone can do, personally and in the age of Instagram, and Apple knows how to use that to sell new phones.
The equivalent of the consumer camera in the corporate world is – or rather, it should be but isn't – AI. It has to be as useful to a company as a camera is to a consumer, which means it has to be completely controlled by the company and as private as each user requires. It has to deliver instantly appreciable results that can be iterated quickly through tools and controls that are precisely tuned to matching need to capability. It must appeal to the heart and mind of a company as much as the camera does to an individual.
What could be so appealing? What does each organization care about? Data. Data about itself, data about markets and sectors, data that helps employees with all the frankly ugly bits of IT like calendaring and getting out of silos, but always the data that is as personal to an organization as a camera roll is to a phone user. That doesn't sound much like the AI being pushed out as part of Windows, because it isn't.
It should be. Microsoft has a huge advantage here because it has an unparalleled ability to understand not just what is normally thought of as corporate data, but the way it's used and how it flows. That would be a lot of AI assistance on the desktop as well as crunchy new services in the datacenter. Good for the ecosystem's refresh cycle, especially if you keep innovating. There's a lot of room to innovate.
The corporate data camera would move Windows beyond its version-based desktop obsession. It will help PC makers who can sell new hardware with actually useful AI acceleration from chip makers who are fiercely competing in a space that's not quite solid enough for comfort.
Focus on innovation? Now there's an idea. ®