Open Letter Signed by Thousands of Authors Urges AI Makers to Stop Stealing Books

More than 8,500 authors of poetry, fiction, and non-fiction signed an open letter criticizing the makers of large language AI models like ChatGPT, Bard, and LLaMa.

Calling out the tech companies for copying their work to train the AI models without seeking their permission or compensating them, the writers urged them to stop stealing their books.

These technologies mimic and regurgitate our language, stories, style, and ideas. Millions of copyrighted books, articles, essays, and poetry provide the food' for AI systems, endless meals for which there has been no bill.Open LetterAI Makers Have Used Their Works Illegally, Claim Writers

Writers signing the letter included notable names such as Margaret Atwood, Nora Roberts, Michael Chabon, and Viet Thanh Nguyen.

The letter pointed out that not only did the tech companies behind the LLMs not seek the writers' permission before using their work, but many of the works were sourced illegally from piracy websites.

Large language models such as GPT-4 can do a fairly good job when asked to adopt the writing style of a certain author when generating a piece of content.

This is because the works of the said authors have been fed into the system during the training process, allowing the LLM to replicate their ingenuity and use it as its own. Naturally, the authors aren't thrilled about this.

It's worth noting that the letter hasn't made any legal threat against the AI companies. Mary Rasenberger, the CEO of the Author's Guild and a signatory of the open letter, said that lawsuits are expensive and take a long time.

However, authors such as Sarah Silverman, Paul Tremblay, and Mona Awad recently filed class-action lawsuits against Meta and OpenAI for training large language models on pirated copies of their books.

Gina Maccoby, a New York-based literary agent, said that she, alongside other agents, has been talking to publishers about including rules in writer's contracts that prohibit the unauthorized use of AI.

What I hear from colleagues is that most publishers are amenable to restricting certain kinds of AI use.Gina Maccoby

In March this year, the Author's Guild updated its model contract to accommodate language addressing the use of AI.

However, Maccoby admitted that even if AI companies are explicitly forbidden from scraping and profiting from literary works as per the authors' contracts, such rules would be difficult to enforce.

She pointed out that there might not be any solid evidence that a certain book was used in the training dataset ingested by an AI program.

Authors' Livelihoods Are Being Affected by AI-Generated Works

Alongside the issue of the unethical and illegal use of literary works in training AI models, the letter also claimed that the large language models are threatening the livelihood of authors.

As of late, publishers are flooded with AI-generated books, and several low-quality AI-generated works entered the Amazon bestseller ranks.

Generative AI threatens to damage our profession by flooding the market with mediocre, machine-written books, stories, and journalism based on our work.Open Letter

The letter warns that the complexities and narrow margins of large-scale publishing allow fewer authors to earn a living from writing and put young writers in an untenable situation.

The letter urges tech companies behind the AI models to obtain the writers' permission before using their works and compensate them fairly for the past and ongoing use of their works in their generative AI programs.

It now remains to be seen if any of the tech giants would apologize and pay the writers whose works they used or if the signatories would eventually take a legal course of action.

The post Open Letter Signed by Thousands of Authors Urges AI Makers to Stop Stealing Books appeared first on The Tech Report.