Diamond marketing group recently promoted COO Alan Macdonald to the newly created role of chief AI officer, tasked with guiding the agency’s use of AI in its everyday work.
Putting Macdonald into the new role is a signal of just how seriously Diamond is responding to the breakneck advances in AI which, while dominating advertising conferences and industry think-pieces in 2023, is only starting to have an impact on the industry.
Putting Macdonald into the new role is a signal of just how seriously Diamond is responding to the breakneck advances in AI which, while dominating advertising conferences and industry think-pieces in 2023, is only starting to have an impact on the industry.
"We aren’t talking about incremental change here. I don’t think our industry fully understands how transformational this will be from an effectiveness standpoint," said agency president Josh Diamond, in announcing Macdonald's new mandate.
"This is about to uproot how we all currently think about going to market. This leadership role will help usher in a new era, not only for Diamond and our clients, but the broader industry as well.”
But aside from his work at Diamond, Macdonald wants to help a lot more people across the industry prepare for what many feel will be an AI revolution in the ad industry.
MacDonald was selected and trained by the international organization VJAL Institute as an AI trainer, and has signed the organization’s pledge to train one million future AI leaders. Outside of weekly workshops for Diamond employees (pictured above), he plans to conduct a minimum of 25 workshops in the next 12 months to bring AI insights to the industry, though he admits that his personal goal is a much larger number.
With 25 years of industry experience, Macdonald returned to academia during the pandemic to pursue AI and machine learning. He now has an MBA in AI from the University of Cumbria, and a certificate in the implications of artificial intelligence for business strategy from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL). He is presently pursuing an MSc in information management and digital transformation from the University of Salford.
The Message's Emma Johnston-Wheeler spoke with MacDonald about his decision to go back to school, how Diamond is using AI now and where it will change the industry for years to come.
EJW: Can you tell me about the decision to return to academia, to learn more about AI, at this stage in your career?
AM: My career has always been in some way connected to data, and one of my catchphrases now is “no data, no AI.” Before Diamond, I worked at Aimia, the holding company for Aeroplan, and we had a 60/ 70 person analytics team who were all focused on machine learning.
This was before 2019 when the open AIs of the world were widely known. Our teams were actually building machine learning algorithms, one line of code at a time, and so even though I was on the marketing side—product development—I got as close to them as I could. We had a lot of first party data, and the ability to be at the table where actionable insights were originating was phenomenal. So I started pursuing machine learning and the broader AI realm.
I haven’t been to school in 25 or 30 years, and getting back into it, I met some of the most credentialed people I’ve ever worked with, or sat in front of.
EJW: How do you intend to elevate the AI knowledge base?
AM: I'm running full-day workshops with our teams at Diamond, every Wednesday. They’re pretty intensive but the focus is hands on keyboards, so it’s about 80% experimentation. People actually leave having built things, and realize that it’s pretty easy because of the tools we now have. They cover generative AI, analytics, natural language, and art.
With the industry at large I'm running these exact same programs, but they're more focused and customized. For example I just did one with a production agency, and I tailored the content to help them generate storyboards. I found AIs that do that, we prompted and we got results and there was a lot of jaw drop moments for them. I'd say the general starting point is knowledge of AI applicability and how seamless it can be transitioned into the workflow.
[However] I'm not trying to re-engineer an entire agency workflow. I'm trying to take micro moments and solve them, whether it be a simple automation that can help somebody not have to sweat how to do something and just get the machine to do it for them. Or in some cases, when there's data files involved, doing data hygiene, and then we'll build a proof of concept that autonomously does something.
I'll give you a quick example: I won't name the client but we built a proof of concept that looked at all past events and based on the benchmarks of the data for all past events, we were able to predict the KPIs for future events of similar size, scale and so forth, and effectively got them more events for equivalent money.
EJW: How do you anticipate AI being used in advertising creative, bot at Diamond and for the industry at large?
AM: The introduction of custom GPTs and advancements in tools like Mid Journey are paving the way for remarkable innovations. These technologies are enabling the creation of more photorealistic images and sophisticated product designs. As we look ahead, the emergence of text-to-video capabilities is a thrilling prospect. This leap forward will not only enhance our concepting process but also revolutionize the creation of explainer videos and onboarding materials. Imagine seamlessly generated videos that bring concepts to life, offering a vivid, immersive experience. This represents a significant shift in how we communicate ideas and tell stories, making the creative process more dynamic and engaging. The potential for these technologies to streamline workflows and boost creative output in the industry is immense.
EJW: Since you're hoping to elevate the AI knowledge base in the industry at large, is there any consideration to preserving a competitive advantage for your own agency?
A: With AI tools becoming more accessible, our edge at Diamond comes from our deep expertise in applying them effectively. The philosophy is 'no data, no AI.' My experience as a certified trainer, working across various industries, has given me a vast perspective on AI applications.
Collaborating with 150 global trainers under an NDA creates a dynamic think tank, enriching our collective knowledge and experience and how use cases get solved. This isn't just about staying ahead; it's about continually pushing the boundaries of what's possible with AI. For Diamond, this means not only having access to cutting-edge tools but also possessing the know-how to wield them in innovative ways. This unique blend of access and expertise gives us a significant competitive advantage, positioning us as leaders in the field and enabling us to deliver unparalleled value to our clients.
EJW: How is Diamond using AI with clients?
AM: Here’s how Diamond is currently using AI and machine learning for clients:
1. Generative AI: We're leveraging AI for category research and summarization. This includes experimenting with creative copy for headlines and call-to-actions, a process that significantly enhances the impact and effectiveness of our campaigns.
2. Image Generation: AI is also playing a crucial role in generating images to help sell concepts. This not only speeds up the concepting process but also adds a layer of innovation and creativity to our work.
3. Natural Language Processing and Analytical AI: We're using NLP for sentiment analysis and research, which informs our content development, particularly in SEO. This approach ensures our SEO strategies are not only effective but also continuously optimized.
4. Marketing Performance Analysis and Reporting: AI tools help us in analyzing marketing performance, thereby providing insightful and detailed reporting. This allows us to make data-driven decisions and strategies for our clients.
5. Automation in Business Case Modelling: Finally, automation through AI is being utilized in business case modelling, streamlining processes and enhancing efficiency in developing compelling business cases for our marketing initiatives.
EJW: Do you have predictions for AI in 2024?
AM: I think in the next six months, we're going to see a lot more customized language models that are unbelievably accurate on a niche subject. A large language model is trained on everything but nothing specifically, so you can ask it anything and it gives you a pretty decent answer back...You get very good responses, but they're sort of generic. What I think is going to come of it is many small language models that are supremely knowledgeable and trained on a specific vertical like law, or health. And then you could go even further into health and look at different verticals within health, and they'd be so supremely knowledgeable and accurate.
EJW: What would you say to people who have anxiety about job loss?
AM: There's a famous tweet I refer to: "AI will not replace you. A person using AI will."
In the early 2000s when we started making websites, [people were asking] "how do you find agencies that know how to make websites?" I feel it's the same kind of history repeating itself. So get on board, use it, and get knowledgeable of it fast. I think to be knowledgeable of the space is sector and industry agnostic.
It shouldn't be a threat, it should be an enabler and and empowering tool to help you take care of mundane tasks, fast. What I always tell people is, I have never used what an AI gave me, I've always taken it and tweaked and augmented and built from it. It means starting at 60% immediately, as opposed to starting at 0% as a human writing. You can use the springboard to get you to get you going.
