AI has ruined Wimbledon…
We can’t let it ruin TV too. Human flaws are needed to create real drama.
Wimbledon didn’t do it all at once. First came the tests. Then a few courts. Then only for challenges. Then, by 2025: Hawk-Eye on every court. No humans left on the baseline.
A decision rooted in accuracy. And in a way, it’s hard to argue: Hawk-Eye doesn’t blink. Doesn’t flinch. Doesn’t miss.
But here’s what it doesn’t do:
It doesn’t hesitate or doubt its own call
Its calls don’t trigger a contentious gasp from the crowd
It doesn’t spark a confrontation at the net
It doesn’t make a player doubt themselves, even when they’re right
Because the line judge wasn’t just a function, they were a character, a part of the story. Yes, occasionally, they got it wrong. But that’s exactly what made them matter and why the drama of Wimbledon was so compelling. Was a player robbed? Did a wrong call turn the tide of the game? This is whats happens in real live to everyone of us, and it’s why we connect with sports. The injustice, the ability to overcome adversity. These are why sports people can become legends.
The same goes for drama. It’s the tiny human flaws - imperceptible at first - that make acting performances compelling. Today, directors can become so enamoured with process and technology that they overlook the very thing that matters: the nuances of human frailty that only actors can capture.
With AI becoming more and more adept at creating incredible imagery, we’re not far away from using it in our visual workflows. Don’t get me wrong, I’m no luddite. AI-generated imagery will be brilliant for the sector. It will allow new content makers to enter the industry. It’s a great equaliser, tearing down barriers that once kept outsiders firmly out. If you weren’t trained on pro kit, you weren’t welcome. Now all that changes.
But. But.
We need caution and a clear understanding of what AI-generated content does and doesn’t do. Yes it can conjure the streets of Victorian London. But are the clothes, vehicles and buildings correct? In the old days, you’d have consultants advising on every aspect of the production - depending on the budget, of course.
AI can’t direct actors
Most importantly though, AI doesn’t direct actors. It’s an algorithm, a piece of code, not a Director. It looks at how actors have behaved in other shows and replicates those emotions but replication isn’t direction and simulation isn’t performance.
AI’s output is only ever an approximation of how it thinks an actor should behave, not how we actually would act in a given environment, and it’s this nuance that matters. We it’s ignored, we can see the effects.
We can already see the results of losing the human element by looking at Wimbledon. The match plays on. The line beeps. The point continues. It’s smoother. More efficient. And somehow, a bit more forgettable, a bit more boring.
And TV’s in danger of doing the same thing
I read that Sky History has begun using AI for its historical reenactments.
Of course they have - they’re smart. They’ve always been nimble, unafraid to experiment with new models and tech. A small channel with big ideas.
Not only that, they’re right to do it: AI reenactments will let smaller indies deliver shows that previously would’ve been unviable. Armour is expensive. Horses even more so. Weather insurance, extras, makeup, location fees… it adds up fast.
As I said before, AI can bring costs down and access up. It can help more people tell stories in a way they never could before. It’s one thing being able to light a glossy MIV, totally another to film a full-scale reenactment with insurance and risk assessments - and it’s that kind of risk which prevented smaller companies being entrusted with bigger work.
But here’s the rub. It’s great that new entrants will enter the sector because of AI. But will they be able to use it properly?
Will that AI-generated soldier ever stumble in the mud, drop his musket, or stand awkwardly in the smoke? Will his eyes dart, unsure, as the scene unfolds?
Because that’s what real soldiers do. Not because it was written that way, but because humans behave in ways that are entire unpredictable. They shake. Adjust. Sweat and in doing so, they become real.
AI doesn’t do real. Maybe it does accurate. But what it really does is clean. Soulless.
It doesn’t flinch when the gun misfires because the gun never misfires unless you tell it to.
It doesn’t mutter something unscripted.
It doesn’t carry the weight of its own nervous energy into a performance.
It just does what it’s told.
And that’s the problem because human made drama doesn’t live in the expected - it lives in the unpredictable.
We think we want polish. But we really want presence.
I’ve met many a young director who obsesses about the latest kit, using cinema grade cameras. Fine, when I was there I got into this stuff too. But actually, thats not the stuff that matters.
Imagine a war film where every explosion is perfectly timed, every face perfectly lit, every movement smooth. It would feel more like a museum diorama than a battlefield.
We’d switch off. Not because it’s not impressive but because it doesn’t feel alive. There is a famous scene in Sam Mendes 1917 when the lead actor is running along a battle-line as men start their attack - people fall over, they stumble, not because they were told to, but because it genuinely happened by mistake. Mendes kept every slip in there and guess what? It’s one of the most authentic, energetic scenes in modern war movie cinematography.
I’m really cognisant that, to some extent, reenactments have always had a faint unreality - we know we’re not really watching reality - thats what archive is for. However recree plays a vital role too. It immerses you in a story, and when it’s done badly it takes you out of the moment. However, the opposite is true too: when something becomes too slick, it’s just as jarring. Perfection also has its own set of risks.
What you aim for is the Goldilocks zone -where it’s just right -and that, I believe, only comes from authenticity, the unexpected, the compelling and messy human performance.
We need the human. The authentic. The unpolished.
It’s why people gravitate to bloopers. To behind-the-scenes footage.
It’s why we still love live broadcasts - even when they go wrong.
It’s why some of the most memorable moments on television were unplanned.
Because in those moments, we see the crack and in the crack, we feel connection.
Social media is built on these moments. Raw, reactive, unfiltered. Last week I wrote about the danger of making your content too beautiful. AI raises the stakes even higher. It’s not just beautiful - it’s airless.
The human touch always wins
Whenever I can, I work with my favourite editor, a lovely chap called Tony Simmons. He’s lightning fast, smart, brilliant. But he also works in what you might call an old-school way.
Back when we still ingested tapes in real time, Tony insisted on watching the whole thing as it came in.
Not for the dialogue. Not for the shot list. But for the mistakes.
He was watching for the camera wobble. The nervous glance at the start of a take. The moment the interviewee shifted, cracked, looked down the lens.
And he would use them.
He’d find a way to put those little fractures in the cut -not hide them. This was long before Netflix documentaries made it a house style. Before the pre-interview silence and mid-interview tears became tropes. Tony already knew what audiences feel, even if they don’t know it: imperfection = truth.
When we see the set-up, the awkward pause, the nerves - we believe it. We connect to it.
Because that’s how humans behave.
AI can’t spot a crack because it doesn’t know what one means
You could feed AI every archive interview ever filmed. Train it on 10 million human faces. Ask it to replicate emotion and it still won’t find Tony’s moment.
Because AI can simulate emotion, but it doesn’t recognise the value of hesitation. It doesn’t instinctively understand that a shift in posture or a muttered “God…” before the first question might be the most important part of the scene.
Editors like Tony are emotional archaeologists.
AI is just a surface painter.
And that’s the real danger. We’re not just replacing people - we’re replacing artisanal skills from the process. The friction, the doubt, the sparks of real emotion. All of it.
So where does this leave us?
We’re not anti-AI. We’re anti-perfection.
Use AI when it opens doors, not when it closes hearts. Use it to expand what’s possible but not to smooth over what’s powerful.
Because real drama comes from risk. From hesitation. From the moments where the actor blinks, the editor cuts late, the truth slips out.
So:
Use AI to create what you couldn’t otherwise afford—but not what you could have felt.
Don’t polish out the pauses. Build in the awkwardness. Let things breathe.
Encourage the crackle, the stumble, the near-miss.
Trust editors who dig for truth, not the clean take.
Remember: your audience isn’t just watching. They’re scanning and if they sense something’s too perfect, they’ll scroll past it.
Because if we lose those moments, we don’t just lose authenticity, we lose attention.
The algorithm might reward frictionless polish but remember, people still connect through mess.
Let AI generate the setting but let humans command the stage.
So if anyone at Wimbledon happens to be reading this - bring back the line judge before it’s too late…