Update: via The Universal Desktop, it’s getting closer.
Before you’ll get too enthusiastic, it’s not like the next Flash Player will get a real-time Voice-to-Text engine. Instead, Adobe is working on tools to automatically transcribe the speech from a video and embed it into that video metadata when it’s published as FLV (Flash Video). Probably not something that you can’t already, painfully, do manually with the current FLV and it’s metadata. The key point here is simplicity.
This can be used, for example, to easily create subtitles for our videos. But, the main goal of this technology is to let search engines index video content and even deep link into a video relevant time.
Personally I’d prefer a native real-time engine inside the player that weights only 50kb, but, this is also nice 🙂