Very interesting developments. OpenAI has already brought us last year the fascination of AI-generated text (and literature) with GPT-2.

In my opinion, AI shouldn’t be considered as the job killer for creatives — since the creatives are already using AI for their works. AI also shouldn’t be seen as just a tool. AI is like a Muse, who inspires the artist with new topics and approaches, unimaginable before. It shows new ways of thinking, even still without ability to think (it is still not A.G.I.). After all it’s up to us how to interpret new ideas.

Visual power (with StyleGAN2 / 3D Ken Burns Effect / 3D Photo Inpainting), Neural Language Processing with GPT-2, Music models like Magenta — and now JukeBox shows us the new multimedia possibilities in art.

Remember Kurt Schwitters, who already back in 1920–40ies used mixed techniques and assemblages combining all available media into one Gesamtkunstwerk. We are now facing a new epoche of creativity.

Sure, DeepFakes bears danger for politics, for our perception, for the financial state of being, for artists as well. We should look positively and enlightened into the future — to learn distinguishing between generative power for intentional fraud or artistic needs.

P.S. Btw we should be aware that some of the lyrics in JukeBox aren’t AI-generated, but taken from existing songs. Fact check is important. But it shouldn’t be the measure for art, like we’ve seen here (it isn’t AI, but it’s very much our topic: what is art? what is real?)