The new 'copyright' war

20 years ago I co-directed a festival called

in Barcelona .

The poster was unrepeatable.

They came the founder of Wikipedia, Jimmy Wales, and the author of the Declaration of Independence of cyberspace John Perry Barlow, founder of the Electronic Frontier Foundation.

Lawrence Lessig, pioneer of cyberlaw with

, came to present his Creative Commons project and explain why

law is renewed every time Mickey Mouse is about to enter the public domain.

The novelist Cory Doctorow, then the star of the most popular blog on the Net, came to explain why access to culture is more than just the right to use an operating system or play a song.

Artists, writers, lawyers and academics came to talk about culture as the process of modifying what exists to create new things.

From transforming an existing refrain, character, program or literary fragment to make something of their own without asking anyone's permission or paying royalties.

We had to protect the public domain and fair use of ideas against the commercial interests of the entertainment industry and rights management companies in a new age marked by internet access.

Two and a half months later, a Harvard student named Mark E. Zuckerberg launched a website called facemash.com from his college room.

Technically, facemash.com had used fragments of a pre-existing work to create something different from the original.

Specifically, he had taken the photos from the files of the students on the Harvard website to compare them with each other and rate their physical attractiveness.

The university accused him of violating the security of the system, stealing copyrighted images and violating the privacy of the students.

Two months later, he launched TheFacebook.com.

In recent years, numerous entrepreneurs have done the same with third-party content, including Cambridge Analytica and Clearview AI.

It's not exactly the same as Afrika Bambaataa taking Kraftwerk

without permission to compose

or MIA building

over a Clash guitar line.

There is no legitimate use in the literal exploitation of the totality of the third-party material for its own purposes.

The question is: which of the two are similar to ChatGPT, Dall-E, Stable Diffusion, DreamUp, MidJourney and the rest of the generative models trained with protected content?

Last Saturday, Getty sued Stability AI for copying and processing 12 million of its images without permission to train Stable Diffusion.

Weeks earlier, three American artists sued Midjourney, Stable Diffusion and DreamUp for using their work without permission to generate images based on their intellectual property.

Last November, OpenAI and Microsoft received a class action lawsuit for "

hacking on an unprecedented scale."

Its Copilot program, a specialized generative code model, has been trained on millions of lines of code from Github, an open programming repository that Microsoft bought in 2018 for $7.5 billion.

None of the plaintiffs rejects the technology.

The artists ask that three C's be guaranteed: credit, consent and compensation.

And this is the part that infuriates me the most: the defendants use our arguments from 20 years ago to deny those rights.

Without them, this artificial intelligence model is an automatic mass plagiarism machine.

That is the new

war .

Subscribe to continue reading

Read without limits

Keep reading

I'm already a subscriber