Mark Pollard: AI can’t compete with a good strategist yet. Here’s what it would need.

Sweathead - Mark Pollard AI

“The work that will differentiate a creative thinker and a creative agency happens next”

By Mark Pollard, CEO and Founder of Sweathead

Have you ever looked at a piece of research, or creative brief and thought, “Yeah but tell me something I haven’t thought of before”?

That’s where AI is right now. It can help you gather information and make sense of it but it won’t see around corners for you. That remains your job. For now.

Research has shown that current AI tools help low performers, or people new to a field, perform at higher levels; but it’s yet to help high performers, and experts, break through in meaningful ways.

This might be one of the reasons that AI usage, specifically ChatGPT, has dropped off in the Northern Hemisphere. According to analytics firm Similarweb, in June from May, traffic to the ChatGPT website decreased by 9.7%. And the time visitors actually spend on the site is decreasing too. It dropped 8.5%, the data suggests.

Tools like ChatGPT are decent at categorising existing knowledge. For the advertising strategist, this means it’s worth asking the tool to list common use cases of a product, perform simple competitor analysis, or write and rewrite single-minded propositions and taglines.

Having said that, ChatGPT can be strategic. Ask it to give you ten ways to solve a problem and it will give you decent starting points, and it will do it faster than your next workshop could.

But the real work and the work that will differentiate a creative thinker, and a creative agency happens next.

For AI to compete with a great strategist, or make a creative team redundant, it would need to do these things:

Become a better writer.

The two main ways this can happen is if you feed it good writing then ask it to mimic that writing style and connect it to a language database.

Researchers at MIT (Massachusetts Institute of Technology) found that memorable  words are monogamous words with few possible meanings and synonyms. For example, the word “pineapple.” Sure, “pineapple” might get used in slang, but, generally, when we see the word, we see the fruit.

Since advertising is about getting into and refreshing memories, a language database of monogamous memorable words would be useful.

AI would need more domain knowledge.

Domain knowledge is the thing that takes years to command. If you’re a stand-up comedian starting out, you’ve only heard so many jokes. But, after twenty years, you’ve probably heard millions more.

For AI to compete with a great strategist, it would need to be fed case studies, effectiveness papers, stand-up comedy, poems and good writing, as well as academic research.

It would need a personality.

Over time, AI can be trained to have personality. Currently, most AI writing is robotic.

Strategists have their own personality, but they develop it and flex it based on their competition and the cultural trends at the time. Ao AI would need to be kept up to speed with these things for it to help.

It would need an audience.

Like comedians, strategists test their thinking on people before it becomes formal. Perhaps different machines (or just the one) would be able to take on different roles, to see if a response is possible.

When a strategist lands an interesting or provocative creative brief, people laugh, or eyes widen, snot travels across the room from a snort. This kind of response can only be tested on a real life audience. 

AI would need morals.

To get AI to a place where it can really help, its moral decision-making needs to be prioritised with pragmatism. 

A great strategist would want the AI tool to be aware of the current moral mood and ethical questions to provide alternative solutions for the strategist to choose from.

Currently, AI is useful. But if you rely too heavily on it, your ideas will be boring and you won’t develop the critical thinking and creativity skills you’ll need in the future when AI starts to take care of the easy stuff.

Mark Pollard is the founder and CEO of Sweathead, a global strategy training company with a community of over 18,000 strategists worldwide and a podcast with over 1.3 million listens.

Pollard’s media career has seen him go from hip-hop journalist to radio host and now, a strategist. His career in strategy started in his native Australia before moving stateside, working for companies including Leo Burnett, Big Spaceship, and McCann. Pollard’s work with the global strategy training company has seen him consult for and train companies like Wall Street Journal, Twitter, The Economist, and agencies around the globe.

Top image: Mark Pollard

To Top