nitter
Greg Brockman
@gdb
6 Apr 2022
So meta:
Mark Chen
@markchen90
6 Apr 2022
"a robot hand painting a self portrait on a canvas" by dalle-2 (
openai.com/dall-e-2/
)
Apr 6, 2022 · 6:20 PM UTC
8
9
3
123
Jameel 🇨🇦
@JJs_playground
6 Apr 2022
Replying to
@gdb
Is the robot hand actually painting something?
1
1
Rafa Jiménez
@rafajim
6 Apr 2022
Replying to
@gdb
Have you asked it to paint itself? i.e. how it imagines itself to be?
2
2
Gašper Beguš
@begusgasper
7 Apr 2022
Replying to
@gdb
DNNs can copy! Not only in the visual domain but also in speech. A similar process in speech:
Gašper Beguš
@begusgasper
10 Nov 2021
Listen to how deep neural networks learn reduplication (copying in speech 🗣 🗣)
direct.mit.edu/tacl/article/…
malcolm
@malcolmtyson
6 Apr 2022
Replying to
@gdb
This is incredible. How long does it take to generate an image?
2
Pradeep
@ppujari
13 Apr 2022
Replying to
@gdb
Beautiful 🤩 thanks
Bernard Piette
@BernardPiette
6 Apr 2022
Replying to
@gdb
The future is now..
cench
@cench
6 Apr 2022
Replying to
@gdb
As a photographer, dall-e 2 caused the same shock I had with gpt2. It’s so good! Can’t imagine what will happen when it reaches gpt3 equivalent in creating images.
2
Leonardo Gutiérrez 🇪🇨
@leonardog27
6 Apr 2022
Replying to
@gdb
😮😮😮 "We, humankind, live in a computer simulation"