AI isn't Exceptional - Humans Are!

Let me explain why the future is bright for the creative industry.

I love the potential that generative AI has, and I'm developing shows with it almost everyday, but while generative AI is becoming increasingly impressive it is actually only able to deliver 'good enough' results and it has creatives across the industry either scared or fuming.

I've seen so many people, especially on LinkedIn it has to be said, who wax lyrical about what the commercial output they've created with AI but yet if I had filmed it with a camera I'd have asked for it to be re-shot for fear that the commissioner would never work with me again.

Every time I have used AI I have had to accept that it isn't quite what I wanted to either say or create. When it comes to LLM's I've had to manually re-write a large proportion of the output. That's why AI will not replace humans in the industry.

The Oscars does not celebrate 'good enough' it celebrates the great and the exceptional.

And that is not achievable with AI. AI requires humans to work with it, in order to perfect it and that will always be the case, regardless of what others will say.

There will never be a point where humans will not be involved in every step of the creative process at some level. Certain sectors of the industry may need to change their skillsets fairly dramatically but humans will have to be involved in the final output.

Creatives need to stop thinking of AI as something that will autonomously take their jobs. It's passive and defeatist. Creative people are inherently creative and will find ways of staying relevant to the industry. I don't know of any commissioning executives or executive producers that even want to remove human input.

Creatives should recognise the value that they bring to a production. Ai can produce 'a result' with a given prompt, but it doesn't know why it produces it. It can't tweak a result based on creative 'knowledge'. It doesn't understand motivation, be that motivation for an actor or for a camera movement. The 'Hallway Scene' in Taxi Driver will stick with me as one of the most telling camera moves in cinema because of what it said about the scene. AI would never have been that intelligent.

AI is replicative, not innovative. Creatives are innovative, especially when it comes to emerging technology. It is estimated that today 60% of the TV industry work in jobs didn't exist 40 years ago.

When you couple this with what is an emerging trend with Generative AI, in that both generative videos and images seem to be converging to create similar results.

This is fantastic for creatives. This shows that AI needs human input in order to create different results.

Let's not kid ourselves, there will be a place for AI driven content. Where there are currently two main genres, Live Action & Animation, I think there will be a third genre created - AI generated content. In this genre, the level of human creative involvement will minimal, maybe limited to just one or two creatives. I suspect that this will exist primarily on social channels, where policy & legislation is harder to enforce.

This is not to say that elements of generative AI will appear in mainstream production, it will, but it will be governed by policy that is currently being written by the industry bodies. On this, it may feel as if the industry is slow to react but in reality the tech is moving so fast that the policymakers are actually having to move much quicker than normal to keep up.

But policy has to be right and not just right now.

But looking at generative AI as individual tools is not really where things get spicy for a creative mind.

The opportunity for AI in the creative space is not with the tools when used in isolation, it's what happens when you try to combine them.

How can the combined tools help you do something or show something that could never have been done before? How can audiences see things that they couldn't have done before? What happens if you use Open AI's Sora with Poly.cam 's 'Splat', for instance? Does that allow you to work in virtual production in a different, more cost effect way?

I'm not an expert in any of those specific bits of software, so as a producer I have no interest in spending my time learning all of that but I sure as hell will hire someone who is an expert.

That's a job created by AI, not one lost by it.

There are so many tools that can be used for pre-vis that couldn't before. Show producers how you can save them time on set by using innovation. Time is money. I will hire you to because you can do that. This works for shot choices, lighting, audio mixes, editing - all of it. Make yourself valuable. I have had a number of meetings just this week alone where we've discussed how AI can streamline our productions but, and this is really important, at the end of those conversations we've all asked -

"Do we want to do that or do we want to keep people employed?"

And the answer is the same, we all want to keep people employed, so we opt for a hybrid model and choose to work with people who can give us the best results using the best technology available.

AI alone offers us 'good enough', humans offer us exceptional.

The best producers want to create exceptional, so help them do that. I sure as hell won't work with anyone who won't engage in the conversation. The industry has moved on so much that anyone who drags their feet over it suddenly feels like a dinosaur.

AI allows for humans to create exceptional content more cost effectively.

This means that execs can take more risks and potentially get more return.

This means that more content can be made.

This means that more creatives can rise through the ranks more quickly making themselves more valuable.

Runners can become producers, camera operators can become highly skilled DOP's and VFX artists.

The opportunities are out there, you've just got to look at things differently. I understand the concern, I really do - the industry is dire enough as it is with the threat of an AI taking your job but in reality it can't.

I've been working with AI for nearly two years and the other day I had to write a document for a film that could result in the largest budget I've ever worked with. Normally I create a first draft of any document with a well known LLM and tweak it from there, but this time I couldn't trust it. The document didn't need to be 'good enough', it needed to be exceptional. So I wrote it myself and it took 2 days. I'm not sure if I reached exceptional but I knew it was the best that I could do and I was proud of it. I know every word of that document and I know why it is there. I could talk to anyone passionately about the film I want to make because all those words are mine.

And that is the difference between AI and Me.

AI can only ever provide 'good enough' and I am always striving for exceptional. Sometimes that means I can work with AI and sometimes it means I can't.

Just because an AI option exists and looks innovative and shiny, would you trust it to deliver output you'd stamp your reputation on or would it just be 'ok'?

Ai on it's own isn't exceptional. But what humans can do with it might just be.