I work on the cutting edge of animation - the conversation around AI misses one crucial thing

The question posed by Will Smith’s Detective Spooner in 2004’s I, Robot.

Now, over 20 years later, the question has skyrocketed in relevance because of the rise of AI and its presumed threat to creative industries.

That threat felt more tangible than ever when, in the past few weeks, Tilly Norwood, touted as the world’s first AI actress, was unveiled, sparking widespread debate.

But while I understand the fear, I think the conversation around the subject gets simplified into a battle between good artists and evil AI. In reality, the two are much more aligned than people think, and the chance of generative AI smothering human-led creativity is close to zero.

Let me explain. Most of what people fear when they say “AI” is a specific subset:  What’s known as generative. A form of AI that’s trained on a massive amount of prior work and remixes it in response to prompts.

The industry I work in already uses advanced machine learning techniques for defined, technical problems like face and body tracking, segmentation, clean-ups, de-ageing, de-noising and restoration.

These are targeted techniques that reduce workload and cost, so artists can spend more time on the emotionality and character of a shot. Crucially, these tools are not substitutes for direction, performance, or authorship. When you look at a finished sequence, everything that makes it land still comes from people making choices, one frame at a time.

That’s different to the use of Generative AI. As it exists today, built on diffusion and transformer models, it cannot create original work - only synthesise the past.

They can produce quick, plausible echoes of what came before. They cannot independently set a purpose, hold a through-line over a feature runtime, or accept direction the way a cinematographer, editor, or animator does.

Professional work needs intent, precise authorship, continuity from shot to shot, characters you can direct, and reliable iteration when the director asks for changes. It needs determinism, long-form coherence, controllability, provenance, and legal clarity. Generative systems are weak in exactly those places.

I get it. When you see AI clips on social media it can feel novel and scary. In reality, it’s a magic trick; it looks like it's pulling something out of thin air, but it's actually using a set of human-defined parameters to search an existing archive of human creativity.

Next time one comes up on your feed, ask two questions: Is this doing something I have not seen before? Does it make me feel anything? (beyond a general anxiety about the technology). If the honest answer to both is no, that is not a failure of audiences; it is a limitation of the tool.

None of this means the technology is “bad.” It means we should use the right tool for the right job. When machine learning helps a filmmaker on a limited budget capture an image they otherwise could not, that is progress. When it is used to compress schedules by deleting craft, that is not progress, it is a short-term cost decision that lowers the ceiling of the work.

There are, of course, real risks. Some executives will hear “faster” and “cheaper” and try to swap tools for talent. Some will blur the lines between licensing, credit, and consent because a model produced something “close enough.”

If we care about sustainable production, we should set the guardrails in contracts and pipelines now. Cutting waste without cutting craft is the only definition of efficiency that matters.

There is also a real upside. When the market is saturated with synthetic images, the value of authored work becomes clearer. Audiences are savvy; they can track authenticity better than some executives will give them credit for. They return to films and shows that carry a human fingerprint, not because they dislike technology, but because intent and care survive the screen.

To go back to the question on people’s lips: will AI replace filmmakers? Well, not in the form that is driving this conversation.

Generative systems are powerful assistants for ideation and reference, and machine learning will continue to remove friction in post. But premium storytelling requires the utmost specificity.

Premium storytelling is the director asking for the subtlest tweak of a performance, a DP shaping the light to catch it just right, an animator protecting eyelines across a cut, a colourist preserving continuity at extreme resolutions. It is purpose-driven. It does not emerge from statistical remixing.

The fear around this technology is understandable. Some of it comes from hype, some from seeing jobs squeezed already.

My view is simple. Use advanced tools where they extend artists. Be clear about rights, consent, and credit. Measure progress by how much more space we create for genuine creativity, not by how much can be automated.

The question was never whether a robot can write a symphony. The question is whether we still insist on what makes a symphony worth writing. That answer remains human.

____________________

Ilija Brunck is the CEO of pioneering animation studio Woodblock, one of the companies behind the visuals at the Las Vegas sphere. The studio has also delivered campaigns for the BBC’s Euros and collaborated with Apple, Nike, Disney, Netflix, Adidas, and MTV.

LBC Opinion provides a platform for diverse opinions on current affairs and matters of public interest.

The views expressed are those of the authors and do not necessarily reflect the official LBC position.

To contact us email opinion@lbc.co.uk