• ai is dead
  • Posts
  • Electric Dreams: Tate Modern's AI Art Odyssey

Electric Dreams: Tate Modern's AI Art Odyssey

PLUS: AI Art Explosion: The Big Three by the Numbers / Sora’s Latest Showcase

On this day in 1955, Disneyland opened to the public (ai generated image)

Walt Disney was once fired from a newspaper job because his editor felt he "lacked imagination and had no good ideas.

July 17, 2024

  • Electric Dreams: Tate Modern's AI Art Odyssey

  • AI Art Explosion: The Big Three by the Numbers

  • Sora’s Latest Showcase

  • Creative Insights with Eryk Salvaggio

  • The AI Wire

  • Creativity in Words: Stan Lee

  • Tools to Explore

Electric Dreams: Tate Modern's AI Art Odyssey

The "Electric Dreams" exhibition at Tate Modern showcases the long-standing relationship between artists and technology, emphasizing that AI should be seen as an opportunity rather than a threat. Opening on November 28, 2024, the exhibition features over 150 works from 70 artists worldwide, spanning from the 1950s to the 1980s. The show highlights how artists have historically used technology to push creative boundaries, offering a unique perspective on the fusion of art and machines.

Perspectives:

Catherine Wood, Director of Exhibitions and Programmes at Tate Modern: "It’s not a new existential threat for creativity. Humans and artists have been grappling with these questions for a long time so we wanted to give the long view on the social, existential and artistic questions around uses of technology to make art."

Catherine Wood: "Art has never just been about crafting images or making imaginative pictures, all the artists in the show are grappling with their being and the technology is a prosthesis or a tool and how those two things come together is what the show focuses on."

Ai Weiwei, Artist: "Art that could be easily replicated by artificial intelligence is meaningless. If Picasso or Matisse were still alive they would quit their job. It would be just impossible for them to still think [the same way]."

AI Art Explosion: The Big Three by the Numbers

In 2023, AI platforms like Midjourney, DALL-E, and Adobe Firefly have generated an immense number of images, with Adobe Firefly leading the count. This surge in AI-generated content underscores the growing impact of generative AI tools in the creative industry.

Almost 15 billion images in one single year.

Sora’s Latest Showcase

We’re buried under an avalanche of AI clips and disjointed montages, and the pile grows deeper every day. Is this just more digital landfill, or a tiny ray of light?

Creative Insights

Eryk Salvaggio
Artist & Researcher
Cybernetic Forests
Cybernetic Forests Newsletter

Q: Your work focuses on understanding and mitigating the harms of AI. Can you share what initially drew you to this field and how your background as a writer and artist influences your research?

A: I make things to understand and reflect on things. My curiosity about tech motivates me, as an artist, toward emerging tools. But as a writer, my interest in structures of power also directs my attention. What does this tech want me to do, and how can I do something else?

Q: In your research, what are some of the most significant harms of AI you've identified, and what strategies do you propose to mitigate these issues?

A: My work is making legible the ways AI reflects the power of those who build it. In AI images, we see stereotypes that often reflect the worst of training data. There's a lot of nasty stuff in these datasets, and traces can make it into the images these tools generate.

Q: As an artist, how do you integrate your artistic perspective into your research on AI, and can you provide an example of a project where this interdisciplinary approach has been particularly impactful?

A: In 2022 I made a series of AI-generated images of humans kissing that were just grotesque! But it also revealed that OpenAI was blocking images of women kissing - while men kissing was fine. That's a distinct, biased human decision that artists could identify that others might not see.

The AI Wire

OpenAI Defines Human Intelligence Levels in AI Systems: OpenAI has developed a five-level system to categorize human intelligence as replicated by AI, ranging from basic capabilities to complex cognitive functions. This framework aims to create a clearer understanding of AI's progression in mimicking human thought processes and abilities. The new system is designed to guide the development and assessment of AI technologies, ensuring a structured approach to achieving human-like intelligence. Slightly unsettling.
For more details, read the full article on Quartz: Quartz.

LightenDiffusion: Efficient AI for Image Generation: The new open-source project LightenDiffusion offers a lightweight diffusion model for image generation. Developed to provide high-quality results with lower computational demands, this AI model is optimized for use on devices with limited processing power, making advanced image generation more accessible to a broader range of users.
For more details, read the full project on GitHub: GitHub.

Creativity in Words

(ai generated image)

“I used to be embarrassed because I was just a comic book writer while other people were building bridges or going on to medical careers. And then I began to realize: entertainment is one of the most important things in people’s lives. Without it, they might go off the deep end. I feel that if you’re able to entertain people, you’re doing a good thing.”

Stan Lee
Comic Book Writer and Editor

Tools to Explore

  • Gan.ai - Simplify the creation and sharing of personalized videos at scale

  • Inro - Instagram co-pilot for DM marketing

Thanks for reading this issue!

Follow ai is dead on:
TikTok / Instagram / YouTube

Follow me on:
LinkedIn / TikTok / Instagram

If you like this newsletter, please tell a friend and get some free stuff!

Let me know if you’d like to advertise.