[jasonlotito]

> What the hell is that part about kids and breakfast?! "Could be used to monitor and train kinds to use the right implements to eat breakfast". That is just messed up.

Messed up? You're just looking at what it's doing, not how it can be applied. The ability to monitor a child with special needs with this sort of application is pretty amazing. Suddenly, this information can be transmitted and used to assist with therapy. Knowing how often a child uses a spoon/fork rather than his hands due to real data rather than parents recalling is powerful.

It also provides powerful feedback. A system setup to remind them child if they stop using their fork/spoon by showing a picture to remind them. This is already used in training. Unfortunately, it can't be automated. Tools like this would allow for that, and the potential is staggering. We are clearly years away, but this has the great potential to really help people.

[harrisonp]

That is a really great point! I was really only appalled by the specific application of audio feedback used in the demo. And as such I blatantly discarded the whole idea of feedback systems. But when I think about it, that area is huge. I really want one of those sensor plates with an arduino interface!

[jasonlotito]

> And as such I blatantly discarded the whole idea of feedback systems.

That happens far too often. You have to disconnect yourself from only looking at one element though. Generally, you have an interaction taking place in the form of Event > Action. Something occurs, and then something happens. In this case, the method of eating is the event. The sound happens to be the action. However, you can always replace the Action with anything else (you can also replace the event with anything else.

Basically, this is how you should approach demos. Not as the end, but as the possibilities.