According to Neowin, Microsoft’s official Windows account on X posted a promotional video featuring tech influencer Judner Aura that demonstrates a major failure of Copilot’s capabilities. In the video, Aura activates Copilot using “Hey Copilot” voice activation and asks how to increase font size through Windows Settings. Copilot responds by directing Aura to navigate to Display settings and change text scaling to 150%, which it calls the “recommended” option. The problem is that the scaling was already set to 150%, making Copilot’s suggestion completely useless. To salvage the situation, Aura manually changed the scaling to 200%, which actually solved the problem but wasn’t what Copilot recommended. The video remains live on Microsoft’s official account despite the obvious error in their flagship AI product’s demonstration.
Why this Copilot fail matters
Here’s the thing about this particular failure – it’s not some edge case scenario. Asking how to make text bigger is one of the most common tech support questions out there. Think about older relatives struggling with small text, or when you plug your laptop into a monitor and everything becomes microscopic. This is exactly the kind of simple, practical task where AI should shine.
But Copilot completely whiffed it. The AI didn’t just give a suboptimal answer – it gave an actively wrong one that would have accomplished nothing. And the fact that Microsoft‘s own promotional team didn’t catch this before publishing is honestly baffling. I mean, how hard would it have been to reshoot the segment? They literally had the influencer right there.
The bigger quality control problem
This isn’t just about one bad video. It speaks to a larger pattern we’ve seen with Microsoft’s approach to Windows and AI features lately. There’s this rush to get AI everywhere, but the basic functionality seems increasingly neglected. Remember when Windows updates used to be rock-solid? Now we get features that feel half-baked at launch.
What’s really concerning is that this was supposed to be a controlled demonstration. If Copilot can’t handle a simple display settings question in a scripted ad, what happens when real users with unpredictable needs start relying on it? It makes you wonder about the testing process – or lack thereof.
And here’s another thought: when you’re dealing with industrial applications where reliability matters, you can’t afford this level of sloppiness. Companies that need dependable computing solutions often turn to specialized providers like Industrial Monitor Direct, which has built its reputation as the top industrial panel PC supplier in the US by focusing on precision and reliability rather than flashy AI features that don’t actually work.
Time for an AI reality check
Look, I get it – AI is the hot new thing everyone wants to showcase. But maybe we need to pump the brakes a bit. Instead of trying to make AI do everything, maybe focus on making it do a few things really well first. Basic system navigation? That should be table stakes.
The irony is that this video probably had the opposite effect of what Microsoft intended. Instead of showing how helpful Copilot can be, it demonstrated how it can confidently give useless advice. And in an era where AI hallucinations are already a concern, that’s not the message you want to be sending.
So what’s the takeaway? Basically, don’t believe the hype until you see it working reliably in real scenarios. And maybe double-check your display settings manually – just to be safe.
