Cat McGinn, industry commentator and the founder of Liminal and Co and partner in HumAIn Media, joins Darren to discuss the structural problems with how the advertising industry is using artificial intelligence. While AI promises creative abundance and more originality, it is instead delivering creative convergence and more of the same.
They explore why AI acts as a “stochastic parrot,” predicting patterns based on past human content and biases, which ultimately regresses output toward the average rather than generating true ideas.
The conversation delves into the cognitive science behind this phenomenon, referencing Wharton’s System Three model and the dangerous rise of “cognitive surrender,” where high trust in AI leads creatives to adopt faulty outputs with minimal scrutiny. To combat this, they discuss how the best use of AI requires systems thinking, taste, and introducing deliberate productive friction instead of simply offloading the effort of thinking.
Finally, they offer an optimistic reframing of creativity as a wild, divergent problem-solving practice. Instead of asking how to make campaigns slightly faster and cheaper, Cat urges the industry to ask what problems can now be solved that couldn’t before.
You can listen to the podcast here:
EMBED PODCAST HERE
Follow Managing Marketing on Soundcloud, Podbean, TuneIn, Stitcher, Spotify, Apple Podcast and Amazon Podcasts.
INSERT QUOTE HERE
Transcription (Edited):
Darren Woolley:
Hi, I’m Darren Woolley, founder and CEO of Trinity P3 Marketing Management Consultancy. Welcome to Managing Marketing, a weekly podcast where we discuss the issues and opportunities facing marketing, media, and advertising with industry thought leaders and practitioners.
While artificial intelligence promises creative abundance, it is in truth delivering creative convergence. This slide towards homogeneity happens because AI acts as a “stochastic parrot,” predicting patterns based on past human content and regressing output towards an average.
Today, we’re exploring structural problems with how the advertising industry is using AI and discussing the rise of “cognitive surrender,” where high trust in AI leads to adopting faulty outputs with minimal scrutiny. Please welcome the founder of Liminal & Co and industry commentator, Cat McGinn. Welcome, Cat.
Cat McGinn:
Thank you very much; it’s good to be here.
Observations on Creative Disruption
Darren Woolley:
You are in a position to provide informed commentary on this topic because you have been observing it for a number of years. From very early on, I remember you bringing a specific perspective to the issue of AI in the creative industries. What captured your interest and motivated you to take that role? You brought a balanced view that was both positive and wary.
Cat McGinn:
I’ve always been interested in technology and how it shapes culture, society, and behaviour. Watching even nascent forms of generative AI back in 2022, it seemed clear it would pose both a real threat and a massive opportunity for the creative industries. I am inherently an optimist, but I think we need to be wary and enter this huge disruption with our eyes open.
Darren Woolley:
On one side, we have technology companies investing billions—maybe even trillions—in the promise of this technology changing society. They are clearly going to over-promise to sell it. On the other side, some have taken a victim-like perspective, fearing the death of careers. It’s playing into the dichotomies we see in the post-social media world, where two sides are battling it out.
Cat McGinn:
That is always the case with new technology; it shines a bright light on cracks that already existed. These issues have come to the fore because the pressure is on. It is revealing structural problems in how we deliver marketing campaigns, how we pay for professional services, and how we interact with one another. All of it falls under scrutiny with this new technology.
The Paradox of Abundance: Quantity vs. Ideas
Darren Woolley:
Someone said to me recently that technology succeeds when it either takes over work you don’t like doing or allows you to do something you could never do before. Does AI fulfil both of those?
Cat McGinn:
Yes and no; it’s a real paradox. The industry has focused for years on wanting to “do more with less,” and AI offers that promise. However, an abundance of output is not the same as an abundance of ideas. That is a tension point we must be cognisant of. I call this “McGinn’s Law”: technology doesn’t necessarily drive us forward. In truth, a new technology often solves a problem directly created by the previous technology.
Darren Woolley:
It’s interesting that you talk about “doing more with less” because when we struggle to measure performance, we default to measurable things like quantity. It’s as if we think, “I’m not sure if this makes a difference, but surely doing more of it will help”.
Cat McGinn:
We get seduced by the speed of output and fail to engage our full critical faculties. We are mistaking quantity for effectiveness, cut-through, and creative originality.
The Risks of Rapid Adoption and “Cognitive Surrender”
Darren Woolley:
This category is evolving quickly; it feels like there is a new breakthrough every week. Where are we currently on the adoption curve?
Cat McGinn:
Twelve months ago, OpenAI was the unquestioned leader, and now we see the rise of Anthropic. Who knows what the world will look like in another year? This level of upheaval is unprecedented. What concerns me is that we are adopting AI wholesale without thinking through the implications, guardrails, regulation, or ethics.
Darren Woolley:
One of my disappointments is how quickly OpenAI turned to advertising as a revenue source. If you position AI as an adviser, having paid advertising corrupts that trust because it may advise based on whoever pays the most.
Cat McGinn:
Trust is a critical problem. We are blindly trusting machines that are riddled with hallucinations. Large Language Models (LLMs) are probabilistic systems; they give the statistically likely next outcome to a prompt. Sometimes they just make things up with such confidence that we forget to check.
Darren Woolley:
Human beings are cognitively lazy. If a tool seems efficient at advising us, it’s easy to just go with it. This is why AI recommendations have surpassed traditional search; it provides one solution, so you don’t have to read through links and make up your own mind.
Cat McGinn:
It was advertising that “poisoned the well” of search, truthfully. Search became unmanageable because the ecosystem was destroyed by SEO and paid outcomes. Now, we see people outsourcing their thinking, taste, and judgement to AI. The risk is a deluge of “slop” and homogenous outputs.
Originality vs. Statistical Convergence
Darren Woolley:
Some argue AI has democratised content creation, but that isn’t necessarily good if people stop judging the outputs. It is just copying patterns from the past.
Cat McGinn:
A large language model is structurally biased towards what is familiar and culturally dominant. Originality doesn’t live in the mean; it lives at the edges, in nuance and cultural specificity. Using them for creativity is where we start to fall down.
Darren Woolley:
We expect it to come up with something new, but human creativity isn’t created in a vacuum either. It is built on past experiences, but humans have the ability to make a “quantum leap”. I find very few marketers have a working model of what human creativity actually is.
Cat McGinn:
When we trust creative outputs to a model that drives us towards statistical convergence, how do you differentiate your brand? Every agency and marketing team has access to the same tools and thinking. How can we connect with customers in a unique way?
Advertising at its best is more than just executing communication in a “professional” way. Outstanding pieces of effective creative have always involved “zigging while others zag”. We can use AI to get to interesting places, but there is a greater need for discernment and craft. There is also a pipeline crisis:
if junior staff aren’t doing the work, where are the future tastemakers going to be trained?
Systems Thinking and Efficiency
Darren Woolley:
Generative AI allows a single human to be more productive if they use it as a researcher to provide options, rather than as a problem solver.
Cat McGinn:
We will see the rise of “systems thinkers” who understand how to hold the whole ecosystem in their head and create constraint and direction. This might be a massive opportunity for neurodivergent people who naturally think in systems and connections.
Currently, agencies are using AI primarily to “do more with less”. Some claim they can do work for 70% less. This focuses on productivity rather than performance.
Darren Woolley:
It’s hard to measure the value or performance being created, so we go for more “widgets”.
Cat McGinn:
We are falling into the trap of efficiency over effectiveness. This isn’t a new challenge; it highlights structural fissures in the industry. We haven’t been paying for great thinking; we’ve been paying for volume. If our only connection with customers is spamming them at volume, it just comes down to media budget size and how much you can interrupt people. I believe there is still value in creating experiences that change how people think and feel about a brand.
Darren Woolley:
Flooding social media with messaging clearly isn’t working. The key is to stand out and get attention. The Campaign Palace used to say on their letterhead:
“No one buys anything while they’re asleep”. Their philosophy was that advertising must wake people up. Now, platforms like LinkedIn are full of programmatically generated “slop”. If you’re struggling to articulate a point of view, maybe don’t make a post if you need an AI to do it for you.
The Danger of Outsourcing Thought
Cat McGinn:
A new research paper from the Wharton School, Thinking Fast, Slow and Artificial, identifies a third mode of thinking. When you outsource your thinking to AI, you stop being able to distinguish whether it was your idea or the machine’s. We start trusting the AI for the perfect outcome and surrender that “sparky” imaginative process.
The study found that the higher the time pressure and stress, the more likely people are to outsource their brains to AI. This leads to significant mistakes in reasoning and analysis and erodes our ability to think.
Darren Woolley:
It reminds me of the Honda “Cog” ad. The agency took months to develop the idea. They finally turned up at the client’s office with the game Mousetrap to explain the concept of engineering excellence through a chain reaction. You can have an idea in an instant, but is it the best idea for a million-dollar investment? AI comes up with the most likely conclusion, whereas creativity uses cognitive load to create fresh, unexpected views.
Cat McGinn:
Like all great ideas, as soon as you hear them, you wonder why they haven’t always existed. We no longer have the time to “mull and process” thoughts because of constant pressure. The promise of automation should be that we outsource the boring stuff—like timesheets—and claw back time for wild, divergent ideas.
AGI and the Future of Human Agency
Darren Woolley:
Companies are researching Artificial General Intelligence (AGI) and super-intelligence. Will that be a game-changer where human beings are taken out of the loop?
Cat McGinn:
I have a degree of scepticism about AGI. It’s interesting that AI companies claim world-ending technology is within grasp just as they go for an IPO. The founder of Anthropic predicted AGI by 2027, yet until recently, ChatGPT couldn’t spell “strawberry” accurately.
This is the moment for us to focus on human agency. What world do we want to live in? Do we want to just feed the machine, or should the machine serve humanity? These conversations are urgent and necessary.
Darren Woolley:
When the creator no longer has a say over the creation, we have lost control and relevance. However, there are great opportunities. In Hollywood, they are shooting in studios and using AI for backdrops, similar to how Gone with the Wind used hand-painted glass, but still using human actors. People rebel against fully AI-generated performances because they don’t feel human. Using it for backdrops can reduce the cost of a high-quality film by two-thirds, which could expand the amount of work available.
The Need for Authentic Connection
Cat McGinn:
The craft of storytelling and human connection will become more vital than ever. This is a crossroads: do we lean into “more of the same for less,” or do we look for horizon-expanding possibilities?
Darren Woolley:
Technology in the 21st century seems to be having a negative impact on connection. Teenagers feel more alone than previous generations. Social media is not fulfilling the need for human connection.
Cat McGinn:
The rise of AI companions shows a human desire for connection, but people don’t know how to make those connections in the real world. However, I see a movement among Gen Z and Gen Alpha to go outside, “touch grass,” and seek messy, real-world experiences rather than social media perfection.
Darren Woolley:
If tools work against the human condition, there will be a rebellion. Sales emails that are clearly AI-generated feel “stalkerish” and lack authenticity.
Cat McGinn:
This makes the need to be face-to-face all the more important; it becomes the only reliable way to interact. Real-life connection in “meat space” is a positive trajectory to be on.
Advice for Staying Curious
Darren Woolley:
Optimistically, where do you see the opportunities for the industry?
Cat McGinn:
If we stop trying to do what we’ve already done but faster and start using AI to deliver visions we haven’t been able to execute, it takes away constraint. It gives us an incredible chance to do new, innovative things that move the needle. Let’s stop doing 10,000 Facebook updates for the same price and find ways to be creatively distinct.
Darren Woolley:
If you were giving advice, what sources of information would you recommend to avoid feeling deluged?
Cat McGinn:
I love several writers on Substack. Empire of AI by Karen Hao should be required reading. Stay curious and look at creators like Phantom X who are building film and TV work using AI. Try to step out of the anxiety of staying on top of every update. Come back to what is irreducibly human: what makes you feel something and what makes you connect?
Darren Woolley:
Cat McGinn, it’s been great catching up. I’m looking forward to the project you’re working on. Thank you for sharing your perspective.
Cat McGinn:
Thank you so much for having me; it’s been great to talk.
