The test is whether a computer can, in an extended conversation, fool a competent human into thinking it is a competent human being speaking the same language,at least 50% of the time.
That is the completely non-scientific populist version. It is basically worthless as it sets no baseline for the jury.
The scientific version of the test includes one human, one computer and a jury. The goal for both the human and the computer is to convince the jury that they are human.
a flat 2D surface in front of your eyes is *not* like a 3D world when your eye tracks (no matter how close it gets). It does not "curve" the same way. So now you need a tiny, bright, hi-res, portable, low-power FLEXIBLE display too which doesn't distort the image too much. Oh, two of them.
Or you could just use lenses combined with a shader that corrects for the lens distortion.
Umm, the middlemen looks to be the music stream services. All they have to do is distribute a product that someone else has produced to the consumer, and for that they are getting to keep a whole 30-40% of the total sale.
Sorry, but if they can't make a profit on that, then they are doing something seriously wrong, or there are too many companies competing for too small of a market.
"There is such a fine line between genius and stupidity." - David St. Hubbins, "Spinal Tap"