- The rise of generative artificial intelligence has a big problem that could hamper its development.
- Potential copyright infringement of large language models like ChatGPT have led to high-profile lawsuits against OpenAI.
- But according to Grayscale CEO Michael Sonnenshein, crypto technologies may offer a solution.
The promise and rapid development of generative artificial intelligence technologies like ChatGPT and Midjourney could be at risk from lawsuits as companies try to protect copyrighted material from being used as a free training resource for AI chatbots.
Last month, The New York Times leveled a lawsuit against OpenAI and its partial-owner, Microsoft, alleging that the large language model used millions of The New York Times articles to train itself.
For its part, OpenAI recently told the UK Parliament’s House of Lords Communications and Digital Select Committee that it would be “impossible” to develop a leading large language model without using copyrighted material.
“Because copyright today covers virtually every sort of human expression – including blog posts, photographs, forum posts, scraps of software code, and government documents – it would be impossible to train today’s leading AI models without using copyrighted materials,” OpenAI said.
Copyright law shaping the direction and development of new technologies is nothing new, and the same legal framework has largely guided the trajectory of the internet over the past three decades.
But what could be new this time around is a solution to the problem, according to Grayscale CEO Michael Sonnenshein.
Sonnenshein told Business Insider Global Editor-in-Chief Nicholas Carlson earlier this week that blockchain technology could lead to a more fair system that helps track ownership of copyrighted material within large language models and image generation platforms — and fairly compensates the owners of that material.
"As much excitement and fervor as there is about AI, there are also some issues that have come up from AI: bias and deep fakes and issues around authenticity and who actually owns what's being produced. To us, it's just so obvious that you need an irrefutable, immutable technology to marry that with to actually head-on address some of the issues, and that technology is blockchain, which underpins crypto," Sonnenshein said.
It's not just OpenAI that is facing legal challenges to its use of copyrighted material in the development of it AI model. Stability AI, a popular AI image generation platform, was sued by Getty Images last year for copyright infringement.
It's yet another issue that could ultimately be solved by the blockchain, according to Sonnenshein.
"Someone creating an image on Midjourney — who actually owns that? And where does that end up living, and how is it used? And if you could, instead, program some of the outputs of this into either tokens or tie it back to blockchain, then all of a sudden issues like provenance and authenticity and ownership, etc, get resolved really, really quickly," Sonnenshein said.
The blockchain is a digital ledger that allows the transparent sharing of information while making it virtually impossible for the data to be manipulated or hacked. Blockchain is the underlying technology that fuels bitcoin, and it has since transferred over to more traditional areas of the economy like in traditional finance and health care.
The blockchain has been touted as a solution for increasing transparency in other assets as well. Experts have said that "tokenizing" everything from real estate deeds and titles to artwork to stocks could heighten security and trust across markets.
It only makes sense for the blockchain to be integrated in generative AI technologies going forward, according to Sonnenshein.
"I feel like crypto has been kind of growing over the last decade as its own kind of industry. And there haven't really been other big technological leaps that could kind of go alongside it. And now I feel like with crypto, and how much time energy and capital is flowing into AI, you can actually see a very symbiotic relationship between these two and how they can actually grow pretty rapidly together," Sonnenshein said.
That symbiotic relationship should ultimately benefit creators and open the door to transparent compensation, according to Sonnenshein.
"As we continue to exploit these technologies, there are going to be programmable ways to ensure that reproductions and licensing and usage of anything from music to artwork — those creators are properly compensated and credited for what they produce," Sonnenshein said.