James Rogers ︎ VFX Supervisor & Creative Director




Photos and sketches actually by James Rogers, not a machine

Creative or Not, Machines are Learning Your Job


Don’t ever think that your creativity can’t be replaced by an algorithm

In early 2021, I noticed a particular sort of Linked In post had started to appear. Some were just copy-paste memes, others were full blown screeds — but they all aimed to communicate the same thing:

“Machines can’t replace me, because I am creative. Can a machine handle a client changing their mind every 5 minutes? I’d like to see that.”

I probably should have ignored it. Social media in the professional domain can often just be grandstanding news posts; or commonly devolves into “connecting” with thousands of people you will never know, let alone meet or work with. Regardless, all my info is up there. Look at me. Measure my resume.

Without doubt, our relevance, real or perceived, is there for quantifying. But these sites also provide a sort of immediate temperature check. Something that you could get from a workplace, but now—given these work from home times—it is something I miss. So when there is a recurring theme to some posts, it seems like we should pay attention.

I have noticed that machine learning has started to move from buzzword to an actual day-to-day interaction for all of us. Mostly, it is almost transparent to our functioning. It might just be on our smart phone, or the way our email is organised, or categorising every photo on our computer. Sometimes it may be more obviously showing up in the software we use. Generally, however, it’s a tool-like implementation — like Adobe’s content-aware fill in Photoshop. It’s clever, simple to use, it mostly works, it can be super-handy — it’s a nice tool. In fact, I think I first read about it on Linked In.

Anyway, I flicked past these can’t replace me posts and closed the app. Unusually, however, I found myself still thinking about these posts much later on. They had stuck with me. And it really annoyed me.

I realised that they irked me because I just found them so incredibly naive.


If there is anything I have learned about keeping up a career, it is the need for constant maintenance. It’s best to regard yourself as never a complete expert, you are always accruing (and forgetting) knowledge, experience and history. Remain hungry to learn or find something else to do.

Sometimes we can be bluffed, but most people can spot someone who knows what they are talking about. Careers can be built on a solid foundation, or a house of cards. We all make our own choices there. Smug confidence will only get you so far, usually. Assuming you can’t be replaced is probably not a great foundation for a career.

To be clear, I am not suggesting that people do not possess expertise, or shouldn’t have pride in their work. A long career in anything comes of listening and comprehending. Especially to people who know their shit. You can still be that expert and keep learning.

But does possessing expertise and a hunger to learn make us impervious to disruption? Has it ever in history? No. Unfortunately for all of us, even creatives, it has not.

Not long ago, if you called a utility company you would be instantly connected to a human, flesh and blood, who would direct your call. Then there came the number menus (press two for…), and now there are voice assistants, “Hello Mr Rogers, tell me how I can help you today”. People used to be adamant that they wanted to talk to a real human, and these automated systems would never take off. Instead, technology moved inexorably ahead, and perhaps all the while we should have been modelling how our futures were bound to change.

Machine learning offers much potential in terms of tools. It can make tedious elements of our jobs faster, like content-aware fill does. Maybe in the near future, an investigative journalist can employ software to make solid leads out of tenuous connections in minutes instead of months; a doctor could diagnose a condition before more obvious symptoms appear, potentially saving a life; an incapacitated person could “speak” fluently without having to slowly type out all the words; or it’s just going to make your video games run faster. Whichever way, these are all tools that positively help our endeavours. There are dark sides to the technology too, but what I am getting at here is change will happen — good and bad.


Now while we may be employed in a creative field we probably don’t appreciate the perceived roadblocks we can present clients. Who is to say that dealing with creatives isn’t a chore some clients would prefer to do without? Would they use a tool that helped them achieve what they wanted? A kind of content-aware fill for the client? Considering we wouldn’t think twice about using tools that eliminate mundane aspects of our job… well, hell yes, of course a client would.

“Why do I have to keep telling this creative what I want every 5 minutes! It’s so tedious…

If creative services people think that dealing with clients can be difficult, then it follows that those clients would feel the same way. A computer interface rarely argues, and doesn’t complain revision after revision. The issue has always been interpretation. The understanding of the creative brief, and the understanding of how to use skill and tools to help produce that result. But if interpretation is less of a barrier, as so much of machine learning seems to promise, then our process simply becomes intent. Accuracy and skilful interpretation gives way to thousands of effortless, but consistently honed attempts.

There is a clear acceleration in the power and application of machine learning, especially those software tools that that take fuzzy inputs. And by fuzzy inputs, I mean the non-expert vernacular of clients.

Those tools exist for writing (no, this wasn’t written by a machine); they are starting to appear for imagery, from colour grading, to photo manipulation, to industrial 3D modelling, to scene building, and beyond; in fact, it doesn’t take a futurist to see that AI will play a part in all aspects of our creative careers.

Imagine the client that wants their ad to have the same colour as their favourite movie? Or their script to be written like Hemingway? Simple tools for that already exist. And they are so simple, even clients can use them. They just might not realise they are driven by AI.

Machines won’t replace humans anytime soon, but they will replace or significantly change some tasks. I really don’t think it is hyperbole to say that in some cases, those tasks may be everything you currently do at work.

There is no safe harbour. No resting on your laurels. We always need to adapt, regardless. Sometimes, that requires a big leap. And the only way to be ready is to take a long view and be prepared. The good news is that this is not the here and now. There is time. We can change, adapt and carefully retrain in anticipation.

But just because you are creative, never think a machine can’t do your job.

Originally published on Medium



Visual Effects Supervisor & Creative Director

︎︎︎ Email

︎︎︎ IMDB

︎︎︎ LinkedIn

︎︎︎ Medium
What I do

A visual effects supervisor is basically a visual problem solver. I try and recruit ideas from a variety of sources—new, old, borrowed or transposed from other disciplines and industries; curiosity drives me. But ultimately, I am an creative image-maker first and foremost.

I closely collaborate with others and keep a holistic view of the process—delivering creatively, on time, on budget. I excel at logistically complicated projects and team building.