Where do we evolve from here? Using #dalle2 in painting onto video frames to generate objects on the leaf. No photoshop. No photos. All AI. AI was used to generate the voice of David Attenborough. Written with some help from @MissyElliott #AIart #dalle @OpenAI

Jul 22, 2022 · 6:42 PM UTC

31
180
41
760
0
The animation was created by masking out a region of the source video and exporting each frame to dall-e with a transparent hole to fill in
4
3
1
73
0
#dalle2 interprets the text prompt in a variety of ways so each frame had to be selected manually. Here’s a look at many of the variations, some of which didn’t get used
1
3
1
48
0
Some of my favorite rejects
3
6
78
0
Replying to @paultrillo @OpenAI
🔥🔥🔥
1
1
29
The only validation I need. I’m done. 🫠
13
Replying to @paultrillo @MissyElliott @OpenAI
Brilliant! Although this probably isn't doable anymore with you having to pay for every prompt with their new commercial model?
1
1
Yes this does mean doing an animation like this will likely use all 115 credits you receive for the $15 charge
3
Replying to @paultrillo @MissyElliott @OpenAI
What generated the voice?
1
1
Replying to @paultrillo @MissyElliott @OpenAI
WTF 😂
2
Replying to @paultrillo @OpenAI
@MissyElliott wanna collab on some oversized grizzly bears?
1
3
Load more