RZ
Robert Zubek
@rzubek.bsky.social
Making games for longer than seems advisable. Big fan of sim games. Doctor of AI, under the hippocratic oath bound to help all game AIs. Architect at Zynga/T2; earlier at SomaSim, EA, etc. Author: mitpress.mit.edu/9780262043915/ Hello from Chicago!
318 followers218 following237 posts
On one hand, I'm impressed with the results! On the other hand, recovering the value of a (hidden) depth buffer channel seems very on-track with recovering the values of visible channels?
True, it's not a big leap from what you can already infer from the visible channels (e.g. SD can render object overlaps reasonably coherently) so I agree, may not actually add much new info on that front.
RZ
Robert Zubek
@rzubek.bsky.social
Making games for longer than seems advisable. Big fan of sim games. Doctor of AI, under the hippocratic oath bound to help all game AIs. Architect at Zynga/T2; earlier at SomaSim, EA, etc. Author: mitpress.mit.edu/9780262043915/ Hello from Chicago!
318 followers218 following237 posts