Hideo Kojima Is Right: AI Should Sweep the Floor, Not Paint the Canvas

Hideo Kojima, the game designer behind Metal Gear Solid and Death Stranding, has a way of cutting through noise. His suggestion that AI should function as a kind of janitor for creative work — handling tedious, repetitive chores so human artists can focus on actual creation — has sparked a genuine debate across the entertainment industry. It’s a useful metaphor. And it raises a question worth sitting with: where exactly does the line fall between AI as a useful tool and AI as a replacement for the creative act itself?

The Janitor Framing, Explained

Kojima’s point isn’t that AI is useless. It’s that AI belongs in a supporting role. Think of the hours a game developer spends on texture cleanup, bug documentation, or generating placeholder assets. A novelist might spend days reformatting manuscripts or cross-checking continuity. A film editor might burn through rough cuts before any real artistic judgment enters the room. These are the floors that need sweeping — and machine learning-powered tools are genuinely good at sweeping them.

The concern kicks in when the broom gets handed a paintbrush. When AI systems start generating narrative arcs, composing original scores, or designing characters from scratch, the question shifts. Is the creative act still happening? Or is it being simulated?

Where AI Already Earns Its Keep

Across the entertainment industry, AI has embedded itself in ways most audiences never notice. Cloud infrastructure lets studios render massive environments overnight. Mobile app development teams use AI to automate quality assurance across thousands of device configurations — phones and laptops alike — without burning out human testers. Augmented reality and virtual reality experiences depend on real-time AI processing to keep immersive environments stable and responsive.

On the production side, machine learning tools can analyze audience engagement patterns, flag pacing problems in rough cuts, or generate subtitle drafts that human editors then refine. These applications fit neatly with Kojima’s janitor model. AI handles volume and repetition. Humans handle judgment and meaning.

Robotics and automation play a similar role in physical production — animatronic rigs, camera systems, lighting grids. Automation handles precision and consistency. The director still decides what the shot should feel like.

The Murkier Middle Ground

Not every application is so clean. AI-powered writing assistants, image generators, and music composition tools have moved into territory that’s harder to categorize. When a screenwriter uses an AI tool to generate five alternate dialogue options and picks one, who wrote that line? When a composer uses AI to suggest harmonic progressions, who composed the piece?

These aren’t rhetorical questions. They’re live disputes in writers’ rooms, recording studios, and animation houses. The entertainment industry’s labor negotiations have made clear that workers see a real difference between AI handling admin tasks and AI generating content that would otherwise require a human salary.

Blockchain has entered adjacent conversations around creative ownership — particularly in how digital assets and AI-generated content are attributed and monetized. The infrastructure exists to track provenance, but the cultural and legal frameworks around creative credit are still catching up. Cybersecurity concerns add another layer, as proprietary creative assets and training data become targets for theft or misuse.

AI, IoT, and the Quantum Horizon

Looking further out, IoT is beginning to feed real-world behavioral data into creative pipelines. Wearables and connected environments could theoretically inform how interactive narratives respond to individual users in real time. Quantum computing, still largely experimental, holds the potential to process creative variables — storyline branches, generative music states, procedural world-building — at a scale current hardware can’t approach.

Both technologies raise the same question Kojima’s framing does, just at higher resolution. More capability doesn’t resolve the philosophical problem. It sharpens it.

The Boundary Between AI and Human Creativity Is a Conversation, Not a Line

What Kojima’s janitor metaphor does well is resist the binary. The debate in the entertainment industry isn’t really AI versus humans. It’s about what kinds of tasks belong to each, and who gets to decide. A tool that cleans up motion capture data is different from a tool that generates a character’s emotional arc. Both use the same underlying machine learning architecture. The difference is what’s being replaced.

Creative professionals across film, games, music, and publishing broadly accept AI in the first category. The resistance concentrates in the second. That resistance isn’t technophobia — it’s a coherent argument that the act of creation carries value beyond its output, and that value belongs to the person doing the creating.

Kojima’s framing won’t settle the debate. But it gives the industry a useful place to start: figure out what the floors are, and sweep those first.