MAD//Fest: So, it’s great to catch up with one of our show's great friend and a speaker at this year's festival! What have you been up to since the event? 

Henry Coutinho-Mason: The main thing I’ve been up to since MAD/Fest is a new project called VisuAIse Futures with Natalia Talkowska, an amazing illustrator and visual strategist who I’ve known for years and always wanted to work with.

It’s an illustrated guide to 6 “not boring” AI innovation opportunities for marketers, and it actually came out of a conversation at MAD/Fest – so it’s great to catch up with you too! We’d been listening to session after session referencing AI, and as we talked we came to this realisation that we could do something together on the topic that was quite unique. 

There’s a lot of people rushing into the AI space – what makes you feel you have a part to play?  

HCM: Our chat over coffee was about how while there was so much being published about AI, most of it felt like it had been written by an AI. It’s not wrong. But it’s not accessible or actionable. It’s not inspiring. It doesn’t compel you to grab your colleagues or clients and say “we need to DO something with this…!”

While we work in very different ways, Henry in words and slides and Natalia in illustrations, we realised that we share the same superpower – creating content that helps people understand change, and that then inspires them to act on it. 

So, you don’t think this whole AI ‘thing’ is a fad? 

HCM: No. It’s clear that it is already changing what we do on a daily basis – I’m sure that most of your readers use AI regularly, whether that’s drafting things, for creative inspiration, or for feedback. Part of our conversation was about how there is an abundance of AI courses and guides aimed at individuals – “Supercharge your productivity using chat GPT” and the like. Then there's all the big macro reports from people like Accenture and McKinsey about how how Generative AI will add trillions of dollars to the global economy. 

But it seems like there's a huge gap in the middle – content that helps people who work in teams within organisations, and whose roles are focused on creating new products, services and experiences within existing brands. This is a huge part of the economy, and it is exactly the MAD/Fest audience, and they aren’t being well served when it comes to AI and its potential. If that’s where you work, how can you use AI on bigger projects, beyond just your own personal workflows? 

That’s just a few reasons why we feel that this more human, more visual, more accessible way of helping people tell stories about the potential for AI will be very useful.

So what are some of the key insights and observations that you’d like to share with the MAD//Fest audience? 

HCM: Actually, the MAD//Fest audience already heard some some of them! I spoke about Virtual Companions and the idea that people will form emotional bonds with virtual characters in July. And the interesting thing is that I first presented this trend on stage right back at the first MAD/Fest back in 2018, talking about how Apple was recruiting a psychologist into its Siri team, because it could see that people were asking Siri questions that went beyond the purely functional. 

I share this story to remind people of two things. First, thinking about people is more important than thinking about technology. Whenever I I'm thinking about new technologies, not just generative AI, the first question I ask is how can these new technologies be used to serve people’s basic human needs and wants? Because the real innovation opportunities will always lie in doing that. It’s never just about using technology for its own sake. Second, even new things have existing roots. Yes, hardly anyone was talking about generative AI before 2022, but of course AI has been a huge topic for many years. 

Of course the other perspective is that some things do change rapidly. For example, one of the big opportunities we look at in the VisuAIse Futures piece is Authorised Celebrity Deepfakes. Now back in 2017 or 2018, deepfakes were super niche – buried in dark corners of Reddit, and basically focused on nonconsensual porn. You could have never walked into a client’s office and pitched them a deepfake campaign. 

Fast forward to 2023, and deepfakes look very different. Now celebrities are embracing the fact they can have digital doubles – especially if that enables them to earn more without any extra work! And for audiences, the fact that AI has become so good and so ubiquitous, means that people can now instantly create personalised deepfakes – unlocking new creative possibilities. So we see campaigns where people can get personalised messages from Jennifer Lopez (Virgin Voyages) or Lionel Messi (Lays).   

Your sessions at MAD//Fest are always very optimistic. Have you carried this forward into your analysis of AI? What about the risks, from brand to existential, or otherwise?  

HCM: Fundamentally yes. Partly, this is because I don’t feel remotely qualified to talk about the existential risk from AI! But on the more practical risks, like bias and misinformation, then yes we do have more of a point of view. But rather than focusing on the negatives, instead we try to highlight the possible solutions. 

So in the graphic we have a section on AI for inclusion – focused on the creative ways marketers are using AI to call out and tackle bias and exclusion. Whether that’s from a very grassroots perspective, like Malik Afegbua, a Nigerian AI creative who used MidJourney to imagine a fashion show for seniors, to the more practical and brand-led initiatives like Estée Lauder’s Voice-enabled Makeup Assistant (VMA), a free app that scans users’ faces and generates voice instructions to help visually impaired people apply makeup better, or Bumble’s open-sourced Private Detector, a tool that uses AI to automatically identify and blur dick pics, enabling users to choose whether to view them or not.    

As readers will be aware, there are huge issues with AI. But our view is that looking for and sharing solutions (even partial ones) is more useful that just highlighting the risks on their own. 

How do you keep on top of all the developments in AI? 

HCM: It’s obviously tough because there’s so much happening right now, but I’ll share a couple of tips. First, it’s about curating the right information ‘diet’. Subscribing to the right newsletters & following the right people will get you 80% of the way there. Second, it’s about having a point of view. I’m always looking for new innovations – startups, products, campaigns – that illustrate a bigger shift, because I find these are the most useful ways to spark conversations and ideas. That’s not just limited to AI – that’s what I’ve always done in my MAD/Fest sessions too – share recent case studies and invite the audience to ask, “what does this mean for our brand or our client’s brand?”

But it’s definitely a fast moving space. Indeed, even since we published the graphic, we’ve seen new examples of the ideas we discuss in it. For example, we talk about the opportunity to be more transparent with when and where AI is used – Google has launched SynthID, a technology that embeds an invisible, permanent watermark on images generated its text-to-image generators. Similarly, Twilio launched its AI Nutrition Labels, that are meant to act like food ingredient labels and disclose how and where companies’ products use AI.  

And there were things we missed that people have alerted us to. After reading our piece on Authorised Celebrity Deepfakes, a reader pointed us to a brilliant yet controversial VW campaign in Brazil which resurrected Elis Regina, a Brazilian singer who died 40 years ago with her daughter. The ad was brilliant – it saw over 50 million views, but also controversial, as while made with the consent of her family, obviously Elis Regina herself never consented to being in an ad like this. So this is still a very nascent space, which is probably why there aren’t that many AI-driven ads yet. Many brands are still watching, rather than doing. 

What did you not include that you’re kicking yourself about?       

HCM: One of the chapters in my book (which was finished just as ChatGPT was released in late 2022), was Augmented Creativity - about how creatives would soon embrace AI as a tool that used correctly would help expand their abilities, rather than replacing them. And I still very much believe that. And so while it didn’t necessarily fit into our graphic because we wanted to focus more on the project-level opportunities that we felt others weren’t covering well, it’s still a massive part of how most marketers experience AI today. 

Sure enough, one of my favourite tools that’s come out since we published our graphic is TextFX from Google Labs, a word game tool that they released with Lupe Fiasco, the rapper. Give it a go, it’s an amazing product and a lot of fun. 

What’s the reception been, and what’s next?

It definitely feels like we’ve hit a bit of a nerve in people. Actually one of the things that has been most well received is the #madebyhumans hashtag we included on the graphic. I think people do feel a bit threatened, and so the fact that we’re confidently embracing that we’re two humans resonates. 

And while it was a very spontaneous collaboration that we spun up over some quieter summer months, the reception has got us excited about the potential for it as a mini brand. Most interesting for us is whether this is applicable in other domains. How much does the overall concept gel with people? We’re having a bunch of conversations with people in various agencies around doing this for specific sectors, and a household brand about doing one with them internally, around their plans for AI. So watch this space! 

Good stuff – we’ll look forward to seeing where you take it. And perhaps one for MAD//Fest 2024? 

HCM: Thanks and great to speak to you, and yes that would be a lot of fun. In the meantime, if any readers want to follow the project then it’s VisuAIse Futures on Substack.