2 ms·
It is a safer bet that someone will hook a neural net up to a camera on someone's hat, train from reality, then re-derive art from that. Then AIs will use the s
by roenxi 4y ago
It is a safer bet that someone will hook a neural net up to a camera on someone's hat, train from reality, then re-derive art from that. Then AIs will use the same methods as humans to maintain their grounding in reality.
The problem with the current crop of AI is they pretty clearly don't understand that they are taking a 2D projection of a 3D concept. That'll be fixed and there is a good chance it'll be fixed with video footage because that is how humans do it.