Commentary: AI didn’t ruin these films. It made them possible

Commentary: AI didn’t ruin these films. It made them possible Bloomberg Opinion's Catherine Thorbecke went to an AI film festival expecting slop, and left surprised by how much humanity emanated from the screen. TOKYO: I went into an AI film festival in Tokyo last week expecting slop, and left surprised by how much humanity emanated from the screen. My favourite was an 11-minute short from the point of view of a child with a camcorder spending a magical summer with his grandparents. He records his cereal, mermaid sightings and fleeting, sun-drenched images of youth. The winner was an animated story about a man in Paris with a facial disfigurement. He hides behind masks but dreams of finding a dance partner. The biggest crowd-pleaser, The Postman, imagined a future in which a trash-sorting robot longs for the human connection it once found delivering handwritten letters as a postal worker. All three films made me feel something, which itself has become a provocative statement. Even suggesting that an AI-generated movie might be good tends to draw indignation from filmmakers and cinephiles who still see the technology as an insult to creativity (it can be), a mass heist of intellectual property (no doubt), or a looming threat to artists’ livelihoods (all true). The anger is understandable. We’ve all seen the soulless junk that defines much AI-generated content online. But what happens when the same technology lets actual artists tell stories they couldn’t otherwise afford to make? HUMAN VS AI Human composition has never guaranteed artistry. On a recent flight, I flicked through several Hollywood movies made by armies of people and backed by enormous budgets. None seemed to have a soul. They were mostly lazy sequels, with writing that felt more like corporate committee work than an expression of human experience. It left me wondering what’s so sacred about that process that we can’t make room for others to create? Audiences tend to rail against AI-generated art - at least, when they know it’s involved. Earlier this year, an anonymous internet artist posted an image on X, described it as an AI imitation of Claude Monet and asked what made it inferior to the real thing. Commentators blasted the lack of depth, cohesion and sophistication. The image was later revealed to be actual lily pads painted by the French Impressionist. The viral episode reflected what studies have found: People often judge AI-generated art more favorably when its origins are concealed, then downgrade the same or similar work after learning it was involved. Rather than proving machines can be artists, this suggests we judge a work partly by how we feel about how it was made. Winning over audiences will require proving that these films involve far more labour than simply prompting a machine. The filmmakers I met in Tokyo were not reckless AI accelerationists. They were artists, often the first to acknowledge the technology’s limits. Yang Wang, co-director of Between Before and After, a moody breakup story, told me that AI reduced costs and improved efficiency, but had not changed the essence. Filmmakers still have to build relationships between characters and organise shots. Long takes, such as Martin Scorsese’s famous nightclub entry in Goodfellas, remain nearly impossible to pull off with AI, he lamented. Sound, color and digital production all changed filmmaking jobs, he added. Some repetitive work might disappear due to AI. But he believes the technology will also create new roles and empower more people to make movies. Yuuuki, the artist behind The Postman, came from anime, where work can be extraordinarily painstaking. He saw AI as a practical way to manage the workload and finish projects. Human creativity stems from accumulated experiences, something a machine cannot replace, he noted. Japan’s storied animation industry has long faced criticism over poor pay and working conditions. Used well, these tools could help preserve rather than replace the craft. COMPENSATING CREATORS The event was hosted by US-based Runway, the US$5 billion startup that makes video and world models. Chief Executive Officer Cris Valenzuela told me that even in the face of loud critics, sometimes from within, “Every Hollywood studio is using AI, every one of them.” It’s true that while executives still publicly emphasise their commitment to human creativity, they’ve been simultaneously investing in AI talent and tools. That shift has produced a sense of fatalism. Artists must embrace AI or be left behind. I’ve long resisted this rhetoric, especially in art. A technology’s existence does not excuse how it was built - by sweeping up vast troves of creative work, often without permission or remuneration. The industry should find ways to compensate those creators. Without their labour, generative AI as we know it would be useless. It can undoubtedly help with tedious parts of production, and many creatives will adopt tools that let them work faster or stretch their budgets. But most workers want a greater say in this transition. That’s why creative industry unions will be essential in determining how they’re used. The debate is often framed as a battle between artists and Silicon Valley philistines. Ultimately, audiences will decide which stories deserve their time and money, whatever tools made them. Consumers are adept at sniffing out inauthenticity whether AI was involved or not. The films I saw were all short, more a glimpse of potential than an immediate threat to the box office. They felt human because a person decided what mattered enough to share. Machines can’t experience the ache of missing a grandparent or the loneliness of hiding behind a mask. AI can change how stories are made. It can’t change what motivates us to tell the ones we do.