Is AI better than you?

First up, we asked our Executive Creative Director, Troy Farnworth, to give his tuppence or two. Spoiler alert: there's quite a few tuppences!

"What comes out of your brain is only as good as what you feed it."

…a good Creative Director once told me.

They were talking about absorbing culture. Taking walks in places you don’t normally go. Take the long way home they said. Watch strange films. Watch how people interact, how they talk. Read books you think you won’t like etc etc. And it was good advice.

As a creative, your work suddenly becomes more interesting. It’s the diversity of input that makes the output so unique and special (Edward De Bono had a few thoughts on this too). It’s also why diversity within the industry is so important, different life experiences, different opinions, different outputs.

Scraping vs inspiration?

So imagine you can absorb everything. And I mean everything. Or as some are calling it within the AI world, ‘scrape’. I find this word really interesting. A lot of chat is around AI scraping the web, infringing on copyrights of artists. I get the points they are making. But all artists scrape. Musicians scrape. Writers scrape. Find me a writer that’s never read a book? All creativity is based on what we are influenced by. The only perceived difference I can see is AI is more honest about its inspiration.

AI can absorb more than any human can. I think that’s a given. But creativity is not just about having good source material – or inspiration as it were. It’s about what you do with it.

Kill your first thoughts

The next bit of good advice I heard was ‘kill your first thoughts’. They’re always too obvious. Any experienced creative can come up with that. When starting out this is hard to take. Anyone can be trained to write better or art direct better, but are they willing to keep coming back to the table with new ideas? This alone filters out those who have long careers and those who don’t.

AI can do the obvious - it can match your first thoughts. And it can keep going and going. But it does need one thing. A prompt. Or in agency speak: a brief.

Now I’m not trying to point fingers here, but most briefs aren’t brief. They’re not exactly ‘Prompt’ size anyway. Deducing some briefs into a comprehendible paragraph is a skill. A skill AI can also do. Prompt GPT to ‘get to the point’ of some waffle and it does a fair job of making it understandable.

AI can’t work out strategy though, can it?

Strategy is a broad term. Most strategy is based on insight. The tools we use to gather insight these days are lightyears ahead of where they were even 10 years ago. From real time direct sales data to audience insight and measurement tools like Kantar, rational, evidence-based strategy is the norm. It’s pretty rare to hear a strategist say ‘I’ve just plucked this out of the air, but I think it’s a good way to go’.

So you have to ask yourself why wouldn’t AI be good at drawing conclusions and point to a ‘way in’ that should be explored based on its own insight gathering?

So AI can take a client brief. Deduce it to something shorter and more comprehendible. Let’s assume it can suggest a strategy or multiple strategies to explore based on insights. It can then be asked to explore. Headlines. Copy. Be it web copy or body copy for ads, scripts or imagery based on these inputs, it can keep producing this content 24/7.

And it can do all this almost instantly. So we’re screwed. Totally utterly screwed.

It’s like when they said robots are going to make cars. Anyone old enough to remember ‘Hand built by robots’? To be clear, it’s before my time too. But today 99% of cars are made by robots. And nobody cares because as long as the output is good the morality of who made it goes out the window.

But we don’t make cars. I hear people saying things like ‘The best campaigns connect with people on an emotional level, AI can’t do that.’

Let’s be honest here. How many of today’s campaigns connect with people on an emotional level? About 1%? What about the other 99%. Remember that thing I said about cars?

AI needs human input

The thing is cars are not entirely made by robots. And that’s the point. You may have seen gaping holes in my hypothesis above. Along the way it needs human input. It’s just a different kind of input. It’s probably much more based on direction setting rather than doing. Prompting can be seen as setting the brief. But refinement of the first response is like creative direction. Not creative direction as in my job but more like how every creative reviews their work during the process of making it. Sometimes it’s subconsciously, sometimes it’s more obvious like a creative team debating ideas they’ve come up with. So how we work with AI will be the most important thing.

There’s an even better side effect. The average work that seems to plague our industry can suddenly be done by a machine. But quicker and cheaper. So more value will be given to those who are really engaging with their audience on an emotional level. Every agency claims to be doing great work. But we know that’s not the case. And there are different opinions on what people think is good. But if a machine can churn out that pretty predictable campaign in 5 mins it’s going to be hard for an agency to charge 3 months’ worth of time on it.

The cream will rise to the top.

Can AI, Chat GPT, machines, robots, really do it all?

I’ve skimmed over some real challenges AI has to overcome before it can really stake a claim within the creative world. Things like, where does it get its inspiration once it’s put many creative people out of business? Maybe we’ll be employed to feed the machine with content? Let’s not fall down that rabbit hole. Red pill, blue pill anyone? Like I say, let’s not go there.

It also has challenges around shifting trends. For example, I’m pretty confident at some point it would be able to create Punk songs one after another. But who decides when Punk is dead and Musick is now the thing? Can a machine really move from understanding current trends to understand future trends? Now that would be impressive.

As someone said, ‘People underestimate the impact of technology but overestimate the speed in which it changes things.’ I think this will be true of AI. It will have a transformational effect in so many businesses, like general computing has before it. The creative industries have always fired back with “it can’t do what I do”. We are now going to find out.

We can question the morality of AI. We can question the legality of AI. But what we really need to be asking ourselves is this. If AI is such a threat in this short space of time, how creative are you in the first place?

We hear from our heart and soul, longstanding copywriter Colin Montgomery next!