AI creative tools are going to be the biggest impact on creative work flows since the computer itself. we are all going to get amazing visual art, music, games, etc.
101
295
109
2,393
(also worth noting: ten years ago essentially no one predicted this. the common wisdom was that we would have self-driving cars very soon but that creativity was out of reach for AI, maybe forever. predictions made today about how AI is going to develop are likely as bad.)

Jul 21, 2022 · 2:25 AM UTC

53
64
24
998
Replying to @sama
I think this is ahistorical - quite a few people predicted this.
Julian Togelius @togelius
21 Jul 2022
There are whole research communities that have worked on human-AI co-creativity for decades, seeding important ideas and building demonstrator systems that tried to establish paradigms that were far ahead of our technical capabilities at the time. What's changed is scale.
1
1
44
Replying to @sama
FYI, 10-ish years ago Peter Norvig and I did a panel where we both predicted that creative applications of AI would be a key area of application. It was at the Commonwealth Club IIRC. Our contention was that computers are less hamstrung by the creative challenges humans face.
1
33
Replying to @sama
They predicted nukes might set the sky on fire, current predictions are probably equally bad.
1
2
Replying to @sama
When do you predict we’ll see the first 100M stream song composed primarily or entirely by AI?
3
1
3
Replying to @sama
True! I admit, I thought the creative fields would be humanity’s bastion… Hiding is no longer an option for us… we will have to find ways to coexist even while this force surpasses us in all ways…
Replying to @sama @tszzl
One reason for this is that a lot of applications for AI in the real world require very high accuracy as errors could be completely catastrophic (self driving cars, manufacturing lines, big robot arms etc) to take over. Art, media generation, etc does have this constraint.
3
Replying to @sama
"Seven years ago, I was a self-driving skeptic. Not of the technology. Of all the 'experts' promising AVs would be everywhere by 2020. You didn’t need to be Nostradamus to know that was ridiculous. All you needed was some common sense." groundtruthautonomy.com/opin… @GTautonomy
1
1
5
Replying to @sama @tszzl
I think the cause of most AI “misses” is direct extrapolation of progress from limited demos. We saw some progress in NL/reasoning with SHRDLU, and everyone thought AGI would be solved in the 70s. With computer vision in 2010s, so self-driving cars would be done in the decade.
1
3
Replying to @sama
Our current predictions are probably wrong as well. Everyone is expecting advanced AI chat bots to help us in our daily lives, but I think most people are underestimating how quickly the AI will become embodied. Humanoid bots will replace labor much faster than we think.
2
Replying to @sama
But also highlights that “creative” is a squishy and capacious concept
1
Load more