For me, voice control of one’s PC, rolling out today in an Insider Build, is a compelling reason to upgrade to Windows 11. Microsoft purchased Nuance, maker of Dragon Naturally Speaking, which I’ve used for years. A while ago in a quarterly earnings report in the usual conversations that important company officers have with stock analysts, investment advisors, etc., Nadella said it was mainly for the use of voice in health/medical applications (when your hands are full working with patients, the ability to dictate commands or text is very helpful). One can speak words about 3x faster than one can type words. When I’m electronically bill-paying, I use Dragon Naturally Speaking(DNS). It’s great especially for dictating numbers, moving between apps, and executing menu commands much faster than I can switch between mouse and keyboard and move the mouse cursor across the screen and click on something, then perhaps start typing again or use the arrow keys. So Microsoft is going to put all the voice recognition love into Windows 11 but not so much, I do believe, into Windows 10.
Could you use voice recognition to significantly improve the MSFS sim experience - maybe one would like to be able to switch graphical settings on the fly in MSFS during live play without exiting what you’re doing - maybe the way the game has to load, etc., it’s not practical to do so. But certainly the kind of power that goes into a gaming PC is really great for voice recognition, too, which involves a lot of computation, especially if there is noise in the background. Maybe if the only thing you ever do with a particular PC is game, Windows 10 is fine until it’s completely unsupported, but for me, I’d rather take advantage of all the future improvements MS is going to put into Windows 11 and not into Windows 10, starting with Android apps, which should show up soon, and improved voice control of one’s PC. Windows 10 voice recognition doesn’t work so well compared to DNS 15 Professional Individual and DNS is getting long-of-tooth and menu commands don’t work well with a lot of modern apps or the changed Windows 11 interface (but then one can record voice macros). So hopefully with MS’s long-term investment in the HoloLens and AR/VR, it’s also got a long-term investment in making voice control/recognition better and better in Windows and that’s going to continue to come along in Windows 11 but I would bet not in Windows 10, which probably is pretty much feature-complete now. I wonder if there will be any future versions of Dragon, too? The incessant e-mails that I used to get for Nuance products have pretty much stopping coming.
Edit_Update: Where speech recognition meets the cloud: DeepMind tests the limits of large AI language systems with 280-billion-parameter model - The Verge.
The article says that Microsoft’s cloud engine for speech recognition is significantly larger than Google’s. So here’s hoping some of that capability trickles down to us lower-level, lower-paying individual customers both in business, personal, and gaming applications. Google’s conclusion, as reported on in the article, is that “big data” significantly improves the performance and the AI that goes along with voice recognition applications.
Here’s an in-sim application for voice recognition in MSFS. As you fly along, just ask ‘Cortana,’ “What am I seeing over the left wing of my plane right now?” or “Give me a bearing to the nearest airport.” (sorry if you can already to this with ATC or an app or the sim already-not keeping up with everything about MSFS).