Can artificial intelligence be creative… be original? Could AI come up with effective marketing?

Those are the big questions that Bant Breen, founder of reputation management firm Qnary, studies.

He says too many people in business – and marketing in particular – lack a true understanding of what AI is… and what it’s capable of. Not to mention the key ways it’s used in marketing already and the potential it has to impact the industry in the near future.

Bant maintains that too many executives are dismissive of AI and machine learning and its potential to change the face of marketing and business.

We talk about the latest developments in this field, as well as…

  • How a machine might replace a marketing executive
  • Why AI is way beyond chatbots
  • How machine learning and AI are the same… and different
  • The surprising uses of AI in advertising you probably missed
  • And more

Listen now…

Mentioned in This Episode: www.qnary.com

Episode Transcript:

Steffen Horst: Welcome to the Performance Delivered: Insider Secrets for Digital Marketing Success podcast. We talk with marketing and agency executives and learn how they build successful businesses and their personal brand. I’m your host, Steffen Horst. Today we’re going to talk about AI and how agencies and media companies can use it to improve several areas of their business. here to speak with me about the topic is Bant Breen who’s the founder and chairman of Qnary, an award-winning reputation management solution company for professionals, brands and enterprises. Bant has worked for numerous marketing holding companies and held different leadership roles at publicists, Dentsu, and IPG. He was inducted into the Advertising Hall of Achievement in 2010. Bant, great to have you on the show.

Bant Breen: Great to be here.

Steffen Horst: Bant, as mentioned a second ago, you’ve worked for a number of big marketing holding companies helping Fortune 100 companies succeed. How did you start your career in advertising?

Getting Started in Advertising

Bant Breen: So the kind of the real professional story really started with me joining WPP on none of a grandiose training program that they set up called the WPP Fellowship. And it was something that they offered to students, mostly graduate students, coming out of college. They were trying to compete with consultancies in the banks. And it was a great… actually probably the best… training program anyone could dream of they, they dropped you into the various WPP agencies, they own Ogilvy and J. Walter Thompson, and obviously several others, and they dropped me in to one of their agencies, I worked there for a year, in essentially kind of a mid-level role, right off the bat. They then sent me to Asia, I worked in Spain and Italy for them, it was it was a spectacular learning experience, and really kind of gave me a very kind of perspective and integrated marketing perspective, at an early stage of my career.

Steffen: So as an advertising executive, you got to spot new trends and ahead of the masses to be able to offer them to your clients and partners. When did you start to look into AI as a solution that could help agencies and media companies in different areas of their business?

Bant: That’s a great question. You know, Steffen, you and I have crossed paths in our career when I was I was running one of the IPG agencies, Reprise. And, you know, one of the things that we were working on at IPG, and I was part of the team that launched, was creating the one of the first programmatic buying groups called Cadreon and you know, the movement towards programmatic buying was very much one of driving to greater efficiencies, handling, more and more kind of niche complexities, you know, very kind of minute complexities and bringing it all together. And, I think that the programmatic era is obviously still full tilt, I think we’re still kind of learning and building there. But this kind of just a simplistic rush for efficiencies, off of kind of old kind of display media… that certainly run its course. And so that it’s, it’s evolved into now handling all different types of media. And it’s becoming more and more complex. And so I started to get interested in machine learning, really, actually, more, I would say, over the last three or four years, not that long.

 As I I realized that marketing and media was kind of it was going to get so complex, complicated, that an individual on their own, could not manage the level of of complexity and the dimensions of the business. And I also started to think about it in an area that’s a little bit outside of media, which was creativity.

You know, if you look at reports that come out of, let’s say, you know, organizations like the World Economic Forum, you know, back in 2015, you know, they talked about the top 10 skills that were needed for executives to succeed in business. And, you know, the number one skill trait was to be able to be someone who could be a complex problem solver. The second most powerful trait was to be able to coordinate with others. The third most important trait was people management. And way down the list at number 10, in 2015, was creativity. Now, we look at it today in 2019 and 2020: complex problem solving still remains right at the top, but they’re nestled in the third position is creativity. And I think the reason why I think the topic of AI and machine learning and creativity is particularly interesting is that that’s when… when machines start to mess around with creativity, that’s when people are uncomfortable, because that seems to be like a very personal no area, we that we as humans seem to think that we have kind of the lock/hold on. So it’s… it was for me to try to understand that I’m certainly Qnary has developed kind of a machine learning layer and its capabilities. It’s very primitive at this stage, it’s really kind of, we use it today for title tagging of content that comes from kind of online publications. But we’re also exploring essentially machine learning as a, almost like a first draft writer for some of the content that we create.

Steffen: So when you when you talk about creativity, you’re not only talking about the visual creative of ad units and things like that.

Bant: No, no, I mean, I think that probably the stuff that I was initially most interested in was trying to understand. You know, first of all, just like articles. Could an article be put together through machine learning, and then I got… I was very interested, you know, McCann Erickson in Japan, about three years ago now created an ad. It’s actually for Mondelez, for one of the Mondelez brands. And the idea basically, was, they actually briefed in a system like gave it a creative brief, and then asked it to formulate like the idea of what an ad, like a TV ad a script would look like. And, and then they made the ad. And it’s, it’s a crazy ad, you should I’ll definitely send it to you, Steffen. Or, or to our listeners, they can go and look it up. Yeah, they ran a contest between an ad made by a very famous producer, a cinematic producer, and this robot to see, you know, what, what could be made. 

At that stage, or at this stage of AI, a lot of this stuff is either nonsensical, or it’s so targeted and prescriptive, that, you know, the stuff that’s made is either like, exact copy, because the system has been trained to do that. Or, in the case of this particular piece that I’m talking about. It’s just, it just grabbed everything. Like, I think one of the things that must have been briefed in was this is for, you know, millennials, and and you know, for things that millennials like, and so the ad basically just has, you know, rocket ships and giant panda bears and all the things that I guess Japanese millennials like?

Crazy.

Steffen: That sounds like a crazy ad.

Bant: Yeah, yeah, definitely take a look at it.

Steffen: Yeah. You know, before we dive deeper into the topic, I wanted to kind of frame it. So people see to use AI and machine learning interchangeably. But is it really the same? And if not, what is the difference?

Bant: You know, I would say, you know, so I’m working… the reason why I’m particularly interested in it right now is in, in the end of September, I’m submitting my dissertation on the topic of, essentially artificial intelligence and advertising. And so, I’ve been, what I’ve been trying to understand is, you know, where are we today as an industry? And, you know, how are things being utilized today? And one of the things that I looked at was that very question, which is, you know, how do you define things? 

You know, in general terms, you know, the artificial intelligence is, you know, kind of described as, you know, an intelligent machine that acts as a rational agent that perceives this environment around it. That’s a kind of a way of generally outlining kind of artificial intelligence. Machine learning is essentially, I would say, a subset of that. Machine learning is kind of how all that really happens, right? And, and so, yes, it gets kind of used interchangeably. I think that’s fine. To be perfectly honest, for most, most of the things that happen in you know, in our industry. But certainly, when we… when I surveyed about this, and I don’t want to kind of, you know, outline too much of the stuff that we that I have coming out, and at the end of September, but there was vast, vast confusion and very little agreement on what artificial intelligence or machine learning was.

Anyway, so. So I think that as an industry, we have, what we’re still at that stage of AI, for some people is the scary robots that you see in a Terminator movie, or Schwarzenegger movie, to, you know, to the, to AI being a chatbot that you see in Facebook, right? It’s really defined in so many different ways. But I would say kind of, like I see AI as almost like a loose/broad term, that would include almost like, all forms of, of artificial intelligence. You know, kind of, so there’s this whole area of artificial general intelligence, which is like when you actually could create a robot or something that actually truly thinks, then, you know, a lot of the components of machine learning, which is very much geared towards, like, if you train, if you train a computer, and you show them a million pictures of dogs, then it will know the next picture is a dog. 

Steffen: You just said, and I think we talked about leading up to the podcast, that you just finished your research dissertation on how agencies and media companies use AI. And you just mentioned one thing about what you found out. What are the things you found? Are there some glaring information or some snippets that you like, “Wow, I didn’t think that, for example, air wasn’t used that much,” or that?

What Even Is AI?

Bant: Yeah, broadly speaking, I looked at about 12 different areas. One was the idea of a collective definition for AI. As I outlined, there really was absolutely no collective definition. Of the five options we gave, it was literally like 20%, across each one. We’ve certainly surveyed several thousand executives in the marketing industry. The knowledge of AI, across the board for anyone outside of outside of like the areas of being data scientist, or some level of a computer scientist, was extremely limited. Mostly still very much in kind of like the flippant “Oh, yeah, well, we’ll be run by robots someday,” you know, just not not really an understanding of what it what it was.

There was no tremendous understanding of how AI was going to impact the organizational structure of work. And, and that’s kind of was one of the things that I found it kind of interesting that at this stage, you know, AI has been around, actually, there has been machine learning in play in a lot of areas of media and marketing, it is actually already influencing the structure of businesses. And yet, there didn’t seem to be any active planning going on in terms of organizational structure and AI. In terms of jobs, there was a very dismissive attitude that jobs could ever do what a marketing people did. So that machines could do what marketing executives did. There very much was a thought that with repetitive tasks, machines could support, but I’m sorry, not one of, but not not an area where, you know, the people were fearful, they actually felt that the marketing area would be one of the last areas to change.

And, in fact, because 95% of all, data scientists, and programmers that are working in the space, are actually working at Google and Facebook right now. You know, and that’s not an exaggeration. You know, they basically have bought all the talent. You know, strangely enough, you know, our AI, the first areas of AI are, are pretty much being driven by the needs, and the plans of those two companies. And so, strangely enough, those two companies are very focused on media and marketing. So in fact, actually, quite a bit, could you know, it does actually have quite a bit of, you know, an impact of, of jobs.

In terms of like, benchmarking where we were where we are today. There was, you know, some knowledge of things like chatbots, but people didn’t really seem to have great experiences with it. So there was nothing that people were awed by. In terms of like, the quality, like is there great marketing that’s been done with machines? No, there was almost like, a general absence of people knowing, you know, there certainly were obviously some award winners that Cannes and things like the ad that I highlighted from McCann. But those were not perceived as excellent, you know, they were, they were perceived as cute or one-off. 

I would say that you know, outside of the survey, you know, we’re starting to see industries that are kind of trend-making industries, like the fashion industry, where they literally have created, you know, artificial, intelligent influencers, right. So like, these characters that sit on Instagram, and literally are, are like, fictional human beings. And, you know, that’s pretty frickin’ creative, as far as I could say. They’re also, I would say that the early breakthroughs that we’re seeing, you know, there’s some kind of stuff that we’re seeing right now that people are shocked by. Like the news today about the Facebook app that allows them to kind of age yourself or makes yourself look young again, but there was a, you know, the deep fake idea of like Joe Rogan, and kind of having the, you know, the deep fake posts. But in reality, that kind of requires so much… it’s not true AI in any way. It’s, you know, it requires a lot of components, but it, you know, it’s very niche in a way.

In terms of like, creativity, you know, the question was, like, “can AI Be creative?” And, well, I, I certainly don’t want to call marketers Luddites, I would say that the vast majority of people that we surveyed, earlier this year, seemed to really think that creativity was still very much the bastion of humans. So there was not a sense that a true thought could be kind of committed. So the idea that you could train something thing to do a great copy. But, could you train something to be original? Even if you could train something to be original, would it be truly original? Or would it still be trained, and you know, there’s a quite a bit of philosophical writing about this topic.

But as you can imagine, the biggest challenge that people kind of see is that we’re in a weird stage now, because we kind of left from like, like a world where we had ad networks and things like that, and, and then we kind of moved into the programmatic space, and the importance of data became more and more relevant.

That being said, the way that we package data, the way we structure it, cleanse, it, is still quite primitive, in a macro sense.

There’s not a lot of understanding of how… so you know, some data sources need to be updated on a very regular basis, some don’t have to be. There’s a lot of misunderstanding of how to structure the data. And if you don’t structure the data correctly, and if you don’t structure it the data correctly at scale, then you’re really not going to be able to move into machine learning. Like, so, I think that you know, when we live in an era right now, where, you know, in the last 18 months, we’ve seen IPG acquire Aaxiom, we’ve seen Publicis acquire Epsilon, we’ve seen Dentsu acquire Merkle. These are large holding companies that are simply just waking up now to the fact that companies need to be thinking about a data strategy outside of providing, you know, the classic agency lip service to date. And so, there’s a lot of work that has to be done on the back end on the data to make it truly meaningful, and that seems to be a huge challenge. 

I mean, I see sometimes kind of like clever user interfaces for that are using kind of machine learning on social media. And I look at the results, and those results will just be horrible because you just look at the data inputs were poorly thought through. So it, you know, we’re kind of at that funny stage.

There’s tremendous excitement about the space. Like, exuberance, but also I would say that that’s matched with just a level of… of fear. You know, just kind of, oh well. You know, I would say, you know, they say that, like, if you argue with somebody, like about politics, no matter what you’ll end up the conversation will ultimately end up with somebody calling the other person a Nazi at some point, you know, if they’re really angry with each other.

And so you know, I think that what I find with the excitement of AI is that at some point, because people have so little understanding of the space, it ends up in a place of fear and a bleak view of the future. That’s why I think, strangely enough, you know, we have not seen any optimistic views of AI, in art or in entertainment. It almost always ends up badly, almost always. And so, and I think that probably kind of parallels where we are.

There’s a lot of concerns about things like ethics, certainly kind of a, you know, racial bias, gender bias, is something that people are starting to think about. And, you know, there are some very notorious examples of AI systems, there’s one that was built by Microsoft, where essentially was a Twitter bot. And after about a couple of hours, this Twitter bot that had been created by Microsoft was, was a sexist, racist, you know…

Steffen: From the wrong people.

The Impact of AI On The Employment Industry

Bant: And yeah, exactly.

So again, to the point of, you know, AI is as good as the data that you put in it, right. So think we’re kind of like still at that stage. There was not a clear view on the future. So, you know, when I asked a question about how do you see a larger role for AI, there’s a sense that it will help them do their jobs better. You know, it’s a fascinating one. 

I mean, I think probably one of the things that I find the most interesting, and apologies Steffen, because I am talking a lot. But when we think about work, I think the thing that I’m kind of blown away right now is the impact that AI will have on work. And what I mean by that is, you know, you just said, you know, we spoke earlier, you know, you have a newborn, baby, congratulations. And as you watch your child, you know, your child will play. And that play is, in fact, actually probably one of the most complex ways someone can learn. The play is so important for a child, because that is actually how they learn about the, you know, the world around them.

Now, why that’s relevant for what I’m talking about now is what happens if all AI takes all of the kind of jobs that you all the, all the crappy work that you and I had to do to move up in our careers? Right? Like, you know, when I called over to Germany, and was like, Steffen, we gotta knock out, you know, search campaign for H&M or whatever it right, you know, and we’re sitting there writing the copy, and, you know, all that stuff is crap work. And now it can be done much more efficiently with machine learning. But, somehow, the learning that you get from doing those, those tasks is lost. You know, if that makes any sense. You know, so like, you know, so I think that you know, in many respects, AI is great for people that know a lot, right? Because, you know, in many respects, we are the trainers, we are the people that can actually kind of manage it. But it’s kind of horrible for young people because they don’t even realize why because it’s literally going to take all the jobs, that would be starter jobs.

So then you say, well, then they’ll work on more complicated stuff. But I’m not, I’m not convinced that there’s the volume of complicated stuff, yet. I’m not convinced that… I mean, I guess what I’m trying to say is how you evolve your career and learn something is going to dramatically change. You will not start from kind of in the mailroom and move your way up. It’s not going to be like that anymore.

Steffen: Yeah, the probability, I mean, that probably requires the younger generation, and probably also the education institutions to look at where is the where’s the employment market going? What are the skills the future generation need to have, in order to be successful in order to overcome the fact that kind of…

Bant: It’s the nightmare of education. So I teach at a school in Spain, it’s great, it’s the top communications University in Spain, and they’re the best kids, they go run the top agencies and all the broadcasters over there. And I am amazed at how hard it is to train them for the future, you know. Because, if you think about it, just training the people on theory and training them on the basics is really, it takes a lot of time in and of itself, and then you kind of have to be at some point, like, okay, everything we talked, you are totally useless. Here’s how it’s really good to go, you need to, you know, you need to kind of, you know, there are some people that are talking about some really crazy ideas. There’s a guy at MIT, who wrote a paper last year about how he imagines a world where literally, like your, your newborn baby would grow up with like a twin, but like a partner for life in a way, like a machine learning that would learn alongside your child. You know, I think that we’re just going to have to… I would imagine young people that are going to succeed in the short term are going to be the ones that are able to utilize AI and its kind of early generation this 1.0 most effectively, right.

So I was meeting with a young woman yesterday, who is trying to get a new job. And she was asking me kind of, like, where to go and what to focus on. And I was, I said, just that, you know, help people, you know, be really go where no one knows right now. Right? Make sure you are the best.

The same way that I think for in my career. And, you know, I was the first one of the first digital guys. Right? So it was known, in a world where people were like, I don’t know what a computer is. Right? So it’s…. and then the last area I looked at is really, where do people begin? And where do they start their learning? And do they have any training in-house and there’s literally no training programs on this inside companies, inside agencies, today.

There does seem to be… McCann seems to be quite progressive in terms of trying to position themselves as a thought leader in the AI space. They produced a wonderful documentary that was released at at Cannes this year that outlines all the areas of artificial intelligence and how will impact the world. But it wasn’t necessarily focused on how artificial intelligence impacts marketing, it was very much focused on just, you know, everything is going to change. And I would say kind of, like I don’t think we’re going to be in a world in the immediate future where, like, we’re subservient to machines or anything like that.

But, I do think that people will get left behind that don’t understand this stuff. You know, and I think that, you know, this will be real… it’s just really important that we do educate people. I think that probably there will be a level of, you know, there’s a lot of software now that’s been developed now, it’s almost like that middle layer of, it allows you to pull together code that allows people to do specific tasks in machine learning.

And so, it’s getting simpler and simpler. There will be there’s going to be a lot of expectation that people will just know how to do a lot of that stuff. So, yeah, so those are the things I looked at. So I mean, honestly, I think broadly actually, I think a lot of that is is shocking, in a way I was shocked that there wasn’t going to be more commonality in the definition, the arrogance in a way that it wasn’t going to be able to impact do what they did was kind of almost like a wonderful, you know, it’s it’s kind of like people sitting on a beach before a tsunami. You know, like, you know, you’re like you’re all gonna die unless you get off the frickin beach like start running.

Steffen: Yeah. If people want to get in touch want to find out more about you. How can they do that?

Bant: I’m sure you know you can always find me on LinkedIn or Twitter but you know the others you can always just email me at my funny name that’s very short but very hard for people to remember somehow. Which is Bant@Qnary.com So Bant@Qnary.com and I’ll be sure to get back to you.

Steffen: Bant, thank you for joining me on the Performance Delivered podcast and sharing your knowledge about artificial intelligence. Now next episode, you and I will talk about what type of agencies are more likely to adopt AI and why, how agencies use AI to improve processes and deliver better results for their client.

 

Share This