Shall we just rename WindowsAppSDK to WindowsAISDK at this point?

[Jimex]

From version 1.7 to 2.0, there’s barely any meaningful progress on UI components. The roadmap feels frozen—no substantial additions, no real evolution—just a noticeable shift toward AI features. Yes, AI is the future, but this is supposed to be an App SDK. Its primary responsibility should be empowering developers to build modern, polished applications with solid, reliable UI foundations.

Right now, that foundation feels neglected. The existing components are far from developer-friendly—take the XAML Designer as an obvious example—and instead of improving these core tools, Microsoft appears to have sidelined UI development altogether. It sends a clear (and frustrating) message about priorities.

What developers actually need isn’t another layer of AI abstraction. We need a smooth, efficient, and predictable development experience—something on par with the simplicity and productivity that WinForms once delivered. Until that gap is addressed, all the AI in the world won’t make building Windows apps feel any better.

[castorix]

Indeed, I cannot use Windows AI APIs on my PC with Windows 10 22H2.
I know it is old, but >35 years old Win16/32 SDK samples still work...
Although I just read they use WinUI to improve Windows 11 : https://www.windowslatest.com/2026/03/21/microsoft-confirms-windows-11-start-menu-performance-boost-shift-to-winui-from-web-based-components/

[ChaimCL]

Microsoft's business strategy is quite confusing. On one hand, they're pushing Copilot PC, while on the other, they're strictly limiting it to a few platforms, completely ignoring the vast number of desktop users who often have discrete GPUs with far better computing power than underpowered NPUs.

They're failing to push their various new AI-related APIs to a broad user base, and at the same time, WinUI 3 itself lacks investment in essential foundational components, trapping them in a terrible cycle of negative feedback. Over the past year or more, there's also been a lack of communication with the community—no roadmap, not even anything to look forward to.

Given that a few days ago, they promised (under pressure from social media) to bring more components to WinUI 3, it remains to be seen whether they truly recognize that the WinUI 3 development toolkit needs to evolve sufficiently.

"If the development tools provide a seamless and polished enough experience, people will naturally use them"—that's still my take.

[danielgruethling]

This is also what is confusing me the most. The applicable usergroup able to run CoPilot in the way they are pushing it is so small the investment they put into it feels like a huge waste. I don't have a CoPilot+ device nor am I willing to run the mess that is the insider build to have the AI features. I don't know any person that does in my professional and private bubble. While I know quite a lot of people that have issues with WinUI and/or Windows in general in regards to performance and stability.
So having basically no investment put into the large paying userbase while the infinitely small group actually using the AI features gets so much is very confusing for me.