Creative teams had two hours to complete the animation.
With some of my amazing colleagues at @GoogleDeepMind, I helped run an animation experiment at the Google Creative House at Cannes Lions. "You're not creating any AI slop today!" https://www.youtube.com/watch?v=meqdlqoLdiI
Users criticized Google DeepMind's animation experiment at Cannes Lions for generating regurgitated slop whose outputs have questionable legal ties to its inputs.
No Digg Deeper questions have been answered for this story yet.
@douglas_eck @GoogleDeepMind Doug, are these models trained on people’s YouTube videos?
I don’t assume they are, and I hope you will say they’re not. But the fact that Google recently said it has the right to do so has a lot of people concerned.
With some of my amazing colleagues at @GoogleDeepMind, I helped run an animation experiment at the Google Creative House at Cannes Lions. "You're not creating any AI slop today!" https://www.youtube.com/watch?v=meqdlqoLdiI
@douglas_eck @GoogleDeepMind What was your models trained on as Ed Newton-Rex had asked?
@douglas_eck @GoogleDeepMind Slop is not just defined by what the output is. It’s the relationship of the output to what’s inputted. Which you’ll surely be clarifying, given you surely don’t want people questioning the source and legality (or morality) of your model. Slop is regurgitated.
