I'm curious, here. What exactly do you mean by a "vibe" in this sense? When a DJ
senses the mood of the crowd, what is he sensing?
The difference between a DJ, or anybody for that matter who is working with an audience, and this robot is this: Simple observation and reaction is not enough. One must have an Understanding of the audience to relate to them. This is fundamental and taught in every speech 101 class. If you do not know your audience, you are going to have a much rougher time. So a sense of culture is important. Unless this robot was programmed ahead of time with a vast amount of information and possibilities, this wouldn't be a factor in its performance. (I doubt it comes stock with that)
Another important aspect of this is Anticipation. Drawing from your Understanding of the crowd, you need to not only react, but anticipate. This involves prediction, a very complex system when relating to Understanding, and something that good crowd managers posesses on an almost unconscious level. To mimic this ability in a robot would require exceeding complex structures.
Timing is the next critical component. If you are merely watching the crowd and reacting to their reactions you are behind the momentum and doing nothing but playing tennis against a wall, which is exactly what this robot is doing. You have to be there with that beat or sound that you've been building up to, and you have to hit it a moment before the crowd expects it. This is something that does not require as much crowd involvement because here is where you are taking the crowd on a journey. It is, after all, your show. As much as you shift it to tailor it to their reaction, they are there to listen to you perform.
So, the points you brought up about sensing metrics, are valid, but they are only a small part of the equation that becomes what I used the word Vibe to explain. So, since all of the above can be learned, YES a robot could do this. Not now though! Not even in the near future, unless this robot was a supercomputer, designed specifically for the task.
Your second question. Would it be easier to divine the mood of a crowd or an individual on a purely observational stance. That's a good question. I don't know enough about pattern recognition to answer it with facts. It would seem to me though that an individual would be easier for a machine. The amount of noise (being motion, audio, ect) that a crowd creates would be a bit oblique. Crowds tend to display the same characteristics across different bulk emotions. Look at a crowd of angry people from afar. You seem arms waving, maybe signs with messages on them, maybe things getting thrown. Now look at a crowd of people celebrating. You see arms waving, maybe signs with messages on them, maybe things getting thrown. :)
Ioa Aqualine Petra'ka
[ Parent ]