Copilot+ is a part of the big AI wave on PCs, but as it stands, a lot of what it's actually capable of is quite disappointing. Its headlining feature, Recall, has been delayed indefinitely, and the rest of what it can do on-device is cool, but nothing that we haven't ever seen before. In fact, I'd argue that as it stands, Copilot+ is behind a lot of the AI that we can already do on our phones, and is definitely behind what a Google Pixel is capable of.

Google Pixel smartphones can do a lot

Copilot+... not so much

If you think about all the capabilities of your smartphone, chances are that you use a lot of its AI features way more than you use Copilot+ or would use Copilot+. Features like "Hey Google" are already significantly more useful than anything you can get with Copilot+, and even live transcriptions on Copilot+-enabled PCs have been on most modern Android smartphones for quite a while now, complete with built-in translation. On top of that, with Pixels, you'll even get other features like suggested replies and Now Playing, which will identify the music that's playing around you.

Other features, like automatic blurring for your camera, can already be done in software and have been done for years. It's not even something that you can only do on a smartphone, it's something that has been around for such a long time that to most people, it's not even a new feature that they're getting to use for the first time. Copilot+ isn't really exciting yet, and even the cool demos that Qualcomm and Microsoft have shown off don't really mean much to the average user. Most people don't care about how an NPU can isolate different tracks in music to separate stems, even if it is cool from a technological standpoint.

To that end, what about Copilot+ is exactly special? What can it do that our smartphones can't? There's talk of on-device generative AI, but even Google and Samsung are starting to toy with Gemini Nano on smartphones now. Plus, do people really care about on-device generative AI? It's certainly interesting, but again, it's not offering useful features like automatic music detection, suggested replies for texts, or even features like Hold for Me and Live Translate.

It's early days for Copilot+

Right now, there's not much special

The problem with AI features like these is that many of them are low-intensity and can run on weaker hardware, which is exactly what's happening with Copilot+. Aside from Recall, all of these features are ones we have experience with in one way or another on either our smartphones or our computers. Most of the benefits of an NPU are that it requires less power, but it can still enable unique features when implemented correctly. With AI burnout at risk of growing further, it doesn't help that a regular consumer can look at all the hype around AI PCs and see that the features advertised are already on the devices in their pockets.

Right now, I'm excited about the future of AI and find the technology behind it fascinating. That said, I'm not exactly rushing out to buy a Copilot+ laptop as soon as I can. I'm happy with my M1 Pro MacBook and suspect I will be for years to come. There's nothing special right now that entices me to get a Snapdragon X Elite laptop, even though I think the Windows on Arm revolution is incredibly exciting.

If you want to experience AI on your laptop or PC, you don't need a Copilot+ laptop. In fact, you can run the most powerful AI on any powerful GPU, and that includes LLMs and image generation. Copilot+ might have a cool future ahead of it, but right now, the features that it offers are ones that people have more advanced versions of on their smartphones already. If you have a Pixel especially, while it's weaker than a Snapdragon X Elite SoC overall (obviously), the AI features your smartphone has undoubtedly outclasses the vast majority of what an "AI PC" can do right now.