What do you think of AI image generators? Do you think computers can be original and creative? Let me know if you've tried out any of the models and what you think!
Instagram: @janpancake
Links:
Diffusion models: • Diffusion models explained in 4-difficulty...
Autoregressive models: • Time Series Talk : Autoregressive Model
Opting out of the next Stable Diffusion training: https://arstechnica.com/information-t...
Interesting papers about creativity, credit, and responsibility:
Hertzman 2018: Can Computers Create Art? (https://www.mdpi.com/2076-0752/7/2/18)
Cetinic 2019: Understanding and Creating Art with AI: Review and Outlook (https://arxiv.org/pdf/2102.09109.pdf)
Oppenlaender 2022: The Creativity of Text-to-Image Generation (https://arxiv.org/pdf/2206.02904.pdf)
Timestamps:
0:00 - intro
0:53 - the past, gans
1:55 - the past, style transfer
3:12 - the past, dall-e
4:57 - the past, vqgan-clip
5:44 - the present
6:12 - dall-e 2, diffusion models
8:00 - imagen
8:39 - parti
9:28 - midjourney
9:54 - stable diffusion
10:19 - personalization of models
11:15 - regulations and policies
11:46 - usage and applications
12:48 - are ai models stealing from artists?
13:35 - can ai models be creative?
14:15 - ai art and credit
15:35 - the future
16:06 - closing
Thanks for watching ❤️
xo janpancake