I don't think obsolescence is a good reason to not advance in something, unless it's exactly known when it'll happen and it's very close by.
The video is pretty great. I'm wondering just how much computing will advance not by means of hardware, but by solving problems we didn't even know existed yet.