you are viewing a single comment's thread.

view the rest of the comments →

[–]UbiquitousCultOfSelf 3 insightful - 2 fun3 insightful - 1 fun4 insightful - 2 fun -  (2 children)

I watched this on mute, but if the image matches what they actually did, then That's some masterful b.s.
The playback of the movie in its entirety vs what the mouse allegedly saw is the funniest part.
Great, it's all glitchy like a rodent on crack... you sold me!
But then with their pin prick eyes and seeing the direction of its gaze, yet the whole movie is there and not as one would expect, a silhouette/vignette akin to a flashlight illuminating a single shifting patch of ground (following the mouse's limited field of vision and wanton gaze).

[–]Maggotus[S] 3 insightful - 1 fun3 insightful - 0 fun4 insightful - 1 fun -  (1 child)

I think the video said that they were reading the signals from the mouses brain and using them to figure out what frames of the movie the mouse was watching. Not actually watching through their eyes, but still crazy enough...

[–]UbiquitousCultOfSelf 1 insightful - 2 fun1 insightful - 1 fun2 insightful - 2 fun -  (0 children)

thank you! I didn't really want to watch it with the audio on all over again.
I guess it's a step.
The real test is not knowing the clip and then recreating it from the brainwaves.
Short of that, it's taking a test while having a cheat sheet.
This will be the true test of AI and big data, is that they are cataloguing the brains "reactions" of firing neurons to particular scenes.
If they have never seen the movie or have no idea which one it might be, to have a neuron firing map or dna if you like of every movie and every scene in existence to then matchup the right one, without knowing the answer, that'll be something.