> have high resolution cameras all over the place. Using deepmagic, object-identify certain things and track their 3D position, relative to some arbitrary point in the area under observation over time.

Say you set up all these cameras on a busy street sidewalk, and you make the origin of the scene the center of a spherical ball on top of the subway entrance hand rail's newel post ...or ballard ...whatever, you get the idea. Pick a thing in the scene that the deepmagic can identify, that is definitely permananetly stationary and that can be used to estblish a point (top-northwest corner of this cube, center of this sphere...)

Do what you got to do to establish a a good origin. That's your origin.

Now, locate and track all eyes, nipples, feet, crotches, etc of all the People. Be precise, now.

It seems reasonable to me that either right now (likely) or in the near future at the latest, it will be possible for cams, computers and deepmagic to track, reasonably well, the lines projected by the center-of-vision for any one eyeball, or all of them.

In other words, you should be able tell exactly where everyone is looking at all times;

Not a person with many cameras pointed right at them from three feet away, with their head carefully braced and held still, and maybe lots of "post processing". No, not that.

You should be able to track where everyone in a scene is looking, pretty darned well, in real time, without any other special constraints.

At least in theory, no? -- Just give it to me.

Worth mentioning: "In realtime" is not really a huge win. Feel free post-process all you like.

And so you've got all the butts and tits and coochies and weiners, and also you've got all the Guccis and sempre fis and ripped jeans and green hair and swole bros.

You've got your Che Guevaras and you've got your Deeply Symbolic American Flag Variants. You've got your Maos but you've also got your Calvins (and they pee on things.)

You've got a sea of interesting things about which you can ask: What do "people" look at and under what circumstances?

You could go on and on with such stuff. On and on. I mean, 8K? more like K9. There's no gloomy news about how sensor resolutions are facing a brick wall or something. I'm sure there are plenty of brick walls, but they exist beyond what's needed to do the above. This, I believe.

Imagine if slutty clothing trends had lots of data to feed on!

"Turns out you just need to wear a shirt with your bank balance/vagina on it..."

[Heh. Naw, I'm just kidding!! I'm in complete denial about the glaringly obvious nature of human behavior, just like everyone else. Be it good or bad, rising or falling, Men, for example, are definitely not on average interested in pounding ladies' cooter holes. Definitely not. Ew.]

There's plenty more applications anyway.

Imagine a security company that tracks what people are looking at as they stroll by the same storefront. "Hmm. They stare at the door handle the whole time and they've been by three times today..."

If stores ever come back, imagine the shitshow that kind of data could produce. I bet if the whole store's interior were covered with faint little RGB LED's, you could SUBTLY direct people's eyes, if not bodies all around like so many servile alpaca. "Hur-Dur! New shiny scooter thing!"

Wanna see a movie for free!!? Just sign right here, blah blah blah.

Looking at signs and shit a lot? Would a deephelper be able to determine that you are "lost" or looking for something? Well, you know where they are looking, so when their eyes cross a "display" of some kind that you control, you can show ...whatever works ...a friendly person waiving their arms and doing the "over here" human thing? Anyway, steal their attention with your display and then say, "Turn left here for the helpful info place!" (??--> in international).

Speaking of which, if a "display of some kind" is cheaper all the time and if we want to send info to a particular random stranger, you know where they are looking and you know what you want to say. One could imagine these "displays of some kind" being a tradable, "liquid" resource for which there'd be mutual interest among retailers (whatever, the "eyeball observation network") to trade and trade freely. (I'm thinking "busy street" or something here.)

It's not a thing, it seems, so maybe the tech is not there. They used to need very expensive cameras to make a "movie", say. Maybe now the expensive thing is cams and gear for this? i.e., You need to scale to a million billion dollar project, maybe in Times Square, and have some good, valuable applications ready for the data.

No, this is not something that you can do just yet, but that doesn't make it useless. People used to stand in front of the one store in town that had a TV. It might suck at first; gotta adapt.

[UPDATE: OMG good lipreading, shee. Maybe people will just stop going outside.]