That’s an interesting futorological exploration, thought it misses one important detail: most of this predictions could work this way only if virtual light technology will also be available. And would it be? We don’t know. Neural link will work as an input device letting user to input some data to machine. But it won’t work itself as an output device. Most likely we still be needing devices to get feedback from machine — both audio and visual. Bone-conducting audio implant the first thing that comes in mind, it’s relatively easy to do such a thing, but with the visual info it is harder. Virtual light is something that we seem to be to far away from, so most likely we will need some kind of a display — it could be glasses or handheld like a phone. So, basically, not a lot will change. You’ll still need a battery-powered device to connect to the network (like an iPhone). You still need a display and something for audio output.