Visual and spatial feedback aren't terribly difficult, though - we already have video-game training gyms, from Atari to GTA. If multimodality and feedback are t

Visual and spatial feedback aren't terribly difficult, though - we already have video-game training gyms, from Atari to GTA. If multimodality and feedback are the main barriers to full AGI, I expect we'll be there soon.