Ed Newton-Rex had reached a breaking point.
But there was growing unease about the movements strategy.
Stability was becoming an emerging powerhouse in generativeAI.
The London-basedstartupowns Stable Diffusion, one of the worlds most popular image generators.
But these two systems were taking conflicting paths.
Stable Audio was trained on licensed music.
The model was fed a dataset of over 800,000 files from the stock music library AudioSparx.
Any copyrighted materials had been provided with permission.
Stable Diffusion had gone in a different direction.
The system was trained on billions of images scraped from the web without the consent of creators.
Many were copyrighted materials.
All were taken without payment.
These images had taught the model well.
Diffusions outputs pushed Stability to a valuation of $1bn in a $101mnfunding round last year.
But the system was attracting opposition from artists including Newton-Rex.
40% off TNW Conference!
Stability was far from the only exponent of this method.
Visual arts, written works, music, and even code are now constantly being reworked without consent.
In response, creators and copyright holders have launchednumerouslawsuits.
Theyre angry that their work is being taken, adapted, and monetised without permission or remuneration.
Theyre also worried that their livelihoods are at stake.
Its in the AI industrys interest to make people think that only the big players can do this.
Artists say that generative AI is stealing their work.The companies behind the systems disagree.
Consequently, the company asserted, there was no copyright infringement.
Newton-Rex considers the practice of exploitation.
Last week, heresignedfrom Stability in protest.
The departure doesnt mean that Newton-Rex has quit generative AI.
On the contrary, he plans to continue working in the field, but following a fairer model.
Its not the impossible mission that the GenAI giants might depict.
In fact, its already been accomplished by a range of companies.
Alternatives are available
Newton-Rex has a long history in computational creativity.
After studying music at Cambridge University, he founded Jukedeck, a pioneering AI composer.
The app usedmachine learningto compose original music on demand.
In 2019, it was acquired by TikTok owner Bytedance.
He was tasked with leading the startups audio efforts.
That objective put him at odds with many industry leaders.
GenAI was edging into the mainstream and companies were rushing to ship new systems as quickly as possible.
Scraping content from the web was an attractive shortcut.
It was also demonstrably effective.
At that time, there were still doubts that the licensed datasets were large enough for training state-of-the-art models.
Questions were also raised about the quality of the data.
But both those assumptions are now being disproved.
What we call training data is really human creative output.
Stable Audio provided one source of counter-evidence.
The systems underlying model was trained on licensed music in partnership with the rights holders.
The resulting outputs have earned applause.
Last month, Time named Stable Audio one of thebest inventionsof 2023.
To me, that showed that it can be done.
Indeed, theres now a growing list of companies showing that it can be done.
One is Adobe, which recently released a generative machine-learning model called Firefly.
As this data is provided with permission, its safe for commercial use.
Adobe also stressed that creators whose work is used will qualify for payments.
Another alternative model comes from Getty Images.
Nvidia has also developed GenAI in partnership with copyright holders.
The tech giants Picasso service was trained on images licensed from Getty Images, Shutterstock, and Adobe.
Nvidia said it plans to pay royalties.
Yet startups areshowing that licensing can also be done on a budget.
GenAI for the people
Bria AIhas provided one example.
The company has developed a new commercial open-source model for high-quality image generation.
You might need to get a little inventive.
You certainly have to do some negotiations and be willing to spend the time.
They need to work to get that in the same way they need to work to get any resource.
If theyre willing to do that, GenAI can work in harmony with human artists.
And hopefully, let all of us enjoy the creativity unleashed by both of them.
Story byThomas Macaulay
Thomas is the managing editor of TNW.
He leads our coverage of European tech and oversees our talented team of writers.
Away from work, he e(show all)Thomas is the managing editor of TNW.
He leads our coverage of European tech and oversees our talented team of writers.
Away from work, he enjoys playing chess (badly) and the guitar (even worse).
