Generate videos with Microsoft Foundry: Review

The Python code block wasn't an exercise. You can't run it, you can't edit it, and there's no sandbox. The entire module is a guided tour of screenshots and text descriptions of what you *would* do if you had access to Microsoft Foundry. I get that giving everyone free access to a Sora 2-level model isn't feasible, but calling this an 'intermediate' module for 'AI Engineers' is a stretch.

I clicked through the whole thing in about 30 minutes. It shows you how to deploy a video model, how to write a prompt, and what the Python code would look like. It's all hypothetical. The knowledge check at the end just asks you to recall facts from the text you just read, like 'What parameter controls the video length?' It doesn't test any real understanding.

I was hoping for at least a sandboxed environment with a heavily throttled, lower-quality model, or maybe just some pre-generated assets to work with. Anything to make it hands-on. Instead, it feels like I just read a very detailed press release. If you're completely new to the idea of text-to-video and want to see what the Microsoft UI for it looks like, this is a quick, free way to do that. If you're a student like me trying to build a portfolio or learn actual skills, this isn't it. It's a product demo, not a lab.