An outraged Richard Osman claims that every book he's ever written has been "stolen" as part of a massive copyright-infringing operation by the creators of the Russian database, LibGen. On his podcast, The Rest Is Entertainment, he said that LibGen possessed tens of thousands of published books, including his own, without obtaining permission.
Adding insult to injury, Meta, Facebook's parent company, has, according to Richard, decided to utilise the text from these illicitly obtained books, articles, and scientific papers to train its advanced Artificial Intelligence engine, Llama 3.
Richard expressed his enormous frustration, complaining: "It's hard to know what to do," as it seemed impossible to prevent. He added that an anonymous individual, identified only by the initials "MZ," apparently authorised the use of LibGen's content.
He speculated that the enigmatic "MZ" might be Mark Zuckerberg, Meta's founder and CEO, although he lacked concrete evidence to support this theory.
Richard, the host of BBC's Pointless and House of Games, discussed an exposé by The Atlantic magazine, which revealed that Meta initially explored acquiring the rights to use the content legally.
However, it became clear that the cost would be prohibitively expensive.
"In all these internal communications, it becomes apparent that it is not particularly cheap and it is not particularly easy," Richard explained, highlighting the challenges of obtaining legal permissions.
He added: "In the communications, someone says 'Look, it is really, really important for Meta to get these books ASAP.
"Another person on the email chain said, 'Look, I've spoken to publishers and this seems unreasonably expensive'."
Following this conversation, Richard claims that the controversial decision was made to exploit LibGen for Meta's AI development: "Apparently someone then gave the go ahead to scrape LibGen, which as we say is a deeply, deeply, deeply illegal website."
He continues by revealing that numerous AI firms are employing similar methods to enhance their Large Language Models [LLMs], justifying their haste by pointing out that in less ethical regimes – such as China and Russia – these practices would be commonplace.
"I understand why, for business reasons, it's happening," Richard concedes, yet he insists that it "really looks" like Meta is capitalising on pirated content.
In the realm of literature, he laments, "there doesn't seem to be any redress for that, and there should be."
An evidently frustrated Richard concludes: "[It's] hard to know what to do, you could spend your entire life taking Meta to court. This is the culture we currently live in, where people can basically do whatever they want if they have enough scale and they're brazen enough."
Mark Zuckerberg, boasting an estimated personal fortune of around $216 billion, is ranked by Forbes as the second wealthiest individual globally, surpassed only by Elon Musk. All four of the top individuals on the Forbes list are involved in funding AI research to some extent.
Bill Gates, who once held the title of the world's richest man, forecasts that only three job types will be secure when AI fully develops: computer programmers, biologists, and energy specialists.
With AI systems already "authoring" books based on human-written works, it's clear that Richard's concerns about writers and other creative professions, previously thought to be immune to automation, becoming obsolete are well-founded.
