AI cannot replace creators… and the proof, surprisingly, is books

The asset-class-ification of entertainment has tended to reduce art to content made only to ensnare views, which lends itself to generative AI uptake. However, in a digital economy increasingly looking towards fandom for solutions, this environment has not exactly fostered the development of ideas, stories, and art worth being a fan of. 

Despite this, there are notable exceptions – and these demonstrate that creators should not worry too much about replacement. 

Some of the biggest successes of the past decade or so — Game of Thrones, Bridgerton, Dune, The Witcher — tap into a story world, allowing different POVs, methods of engagement, and points of sale. This is also true of the longest lasting successes: Lord of the Rings, Dungeons and Dragons, Star Wars. Interestingly, the common denominator for most of these franchises is that they are based on books. 

There is a particular irony in this, given there were once fears that TV would kill books altogether. And yet here books are, their audio versions thriving and their storylines underpinning some of the most successful franchises to-date. It seems that no matter how much money you throw at a production, you just can’t beat good writing. 

The reason is not so much about format as it is about foundation. Game of Thrones was arguably one of the most impactful series of its era, but the last few seasons are considered a bungle of tragi-comedic proportions. Notably, it went downhill after they ran out of books to use (and the lead writers got distracted by a contract with Disney on the horizon). 

Star Wars and Dungeons and Dragons are not based on books, but there was so much creative effort put into building out the worlds behind them that many book series emerged to flesh them out. Both official and fan-made, these books examine different eras, points of view, political events, and canonical truths that inform events within the official instalments. In other words, the creator economy is not new, it’s now just uploaded to TikTok instead of doodled in the margins of notebooks.

Strong source material is not the only ingredient, however, and here is where the ‘asset-class-ification’ of these stories comes in. The assumption is often that the IP is inherently valuable; buy it, make a production, and fans will follow. IP is all. By this logic, AI can be used to fill in the gaps on command, without investing time or resources into doing so. Yet it is not only the use of the material by itself but the way it is used – which can be good or bad (or comical, as the 1984 version of Dune will attest) – that seems to make a difference. Or, put another way, good art is a labour of love, not finance and technology. 

Peter Jackson’s original Lord of the Rings films took nearly half a decade to make, and interviews with cast and crew affirm that it was the experience of a lifetime. Jackson was reluctant to make the Hobbit movies, and when he did end up directing the films nearly a decade later, the delivery was comparatively underwhelming, not least because of the over-reliance on CGI. Ian McKlellan reportedly cried of despair on set. The Rings of Power relied on real sets, but its script was lacklustre. Despite preexisting, ardent fandom around Tolkein’s Middle-earth, the show was deemed a disappointment: it has an average 38% audience score on Rotten Tomatoes. For context, the first Hobbit film has 83%, the Fellowship of the Ring sits at 95%, and Open Water – a shark horror from 2003 with a script written in six days and an estimated budget of $500,000 – has a more comparative 33%. 

Of course, the Middle-earth franchise has changed studios, budgets, and directors – but other examples prove the rule. The Witcher has been a runaway success, moving from books, to games, to the current TV show. However, the lead actor Henry Cavill, a huge fan of the original series, is now leaving Netflix’s The Witcher series – reportedly due to disagreements with the showrunners over the portrayal of his character, which hints that fans may not be pleased with the coming seasons. Bridgerton, on the other hand, has stayed faithful to its source material while putting its own spin on it, meaning fans can enjoy reading the books and watching the series at the same time. 

Riot Games has fleshed out its online multiplayer game League of Legends through story building. This includes two other games in the same universe, the TV show Arcane (96% on Rotten Tomatoes), and a plethora of other published content, from video shorts, to interactive maps, to written short stories. All of these flesh out the characters and storylines with particular attention paid to meaning and quality, which gives audiences something to identify with and care about beyond just the game mechanics. 

Indeed, it seems that all the bells and whistles of high-budget productions can be entirely unnecessary. The self-titled ‘BookTok’ community is a huge and devoted group of fans of various book series, who engage far beyond TikTok – from live events, to podcasts, to fan-run accounts creating their own short-form adaptations. 

The most expensive IP can yield poor results if the way it is used is unsatisfactory to fans. Meanwhile, the most unexpected creations can become overnight successes (or at least subculture icons, like The Room) if they are entirely committed acts of creative passion. Books are a strong way to fill out these creative worlds; after all, there are no frills for the quality of the story itself to hide behind. However, this is not exclusive. At the end of the day, it is not just the concepts behind the stories themselves, but rather how they are executed, that makes the difference. 

As generative AI threatens to flesh out ideas without the ‘creator’ needing or understanding the skills required for the creative process, this is worth remembering. It is not just the what but the how; not just the use of any popular iconography but why and when they are used, that set successes apart from high-budget flops… and these are things AI is still a long, long way from being able to replace.