[pseudocoup]

the ai is not reconstructing sight. it’s guessing which frame of a video the mouse is currently watching. based on training data created by recording the lit up brains of other mice watching the same video

[sailfast]

Interesting. In that case what a terribly misleading / borderline irresponsible article. The video from the study to go along with it was also garbage in explaining this, but I admit I probably got led astray by all the talk of the visual cortex and the skippy frames.

Thank you for this clarification / setting me straight.