How to use artificial intelligence in marketing? Marketing conversation with Paul RoetzerMar 18, 2023
Who is Paul Roetzer?
Yes. Paul Roetzer, I’m the founder and CEO of Marketing Institute. I had a marketing agency. I had founded PR 2020 back in 2005 and out of that was marketing. And so it was born and I sold the agency and I focus all my time now on the Marketing Institute and trying to help people understand and AI to marketing.
What is Artificial Intelligence?
Paul Roetzer: The simplest definition that we have found that I do use in the book is the science of Making Machine Smart. So Demis Hassabis is based in London. He has an AI research lab called DeepMind, which was acquired by Google in 2014. Brilliant, brilliant AI researcher. And that's how he defines it. And so we sort of it's the best definition I've come across because what it means if you think about the software and the hardware you use to do your job or function in your daily life, your phone, for example, an iPhone or a Google device, the software and the hardware that make it up. Think of that as the machine. So the machine is just a word, meaning largely the software hardware, and it's giving it intelligence. So rather than the human having to tell it what to do and setting all the rules, the machine can actually learn and evolve and make recommendations and predictions. So it's all the things we use to do. Our jobs becoming smarter around us, basically.
(...) So the way the search engines work is based largely on what is called large language models, which is what enables writing today. So GPT three, which many marketers are playing around with tools that can write copy those are based on. So that model is from Openai. But Google has Lambda as their large language model, their main one. And so those models are changing the way people interact with information, the way they seek information, and the way they create information. So certainly search engines on their own could dramatically change. I actually saw it I think it was this morning about how Tik-Tok could actually become a search engine or Twitter could actually evolve to become a more powerful search engine. So I'm looking at one today. Yeah. So yes, I think the future of search engines is up for grabs. And I mean, Google is insistent on watching everyone else build the future. I'm sure they have their own ideas of what the search is going to look like. But absolutely. I mean, there is already a search today that doesn't exist without AI. So what Google already does is powered by AI. And I think there's been a lot of breakthroughs in the language in particular recently that could again create a new generation of search engines.
What is Generative AI?
Paul Roetzer: And so that is the hot term these days. It's sort of emerged, it feels like in recent weeks. So generative A.I. means the ability of the machine to generate anything. So specifically for marketers, text and images are the two hot items. So being able to go into a Jasper or writer or go Charlie or copy dot A.I. or hybrid and give a text from. I want a paragraph about artificial intelligence for marketers and the machine generates a paragraph. So that is generative A.I. It's a very descriptive term. Dolly, too, is another example you can prompt to give you an image of whatever you wanted it generates Medak came out with Make a video where you can prompt to actually create a video. You're going to see this applied in all kinds of verticals so I can text prompt, and design my office for me in the style of a lost. And it will be it'll build it out and put furniture in it and things like that. So you're going to see this applied not only in marketing but in every industry that you may work in, your clients may work in. Generative AI is going to be a major, major part of the story in the coming years.
If AI is replacing us, what are we, as marketers, still useful for?
Paul Roetzer: It's a moving target, honestly. So when Dolly two came out in April and made image generation widely accessible, I mean by June or July, they'd opened it up for anyone to get it. When you sign up today and start building images, it changed things in my opinion. I mean, it was one of those I've been paying attention to the space for 11 years, closely. And there's been a few moments where I thought, well, that changes everything.
And I felt like DALLe was a moment where everything changed because now it affects designers and illustrators and makes our lives as marketers easier because we don't need designers and illustrators to do some of that stuff or stock photography. We can do it on our own. So I think that it's moving in a really interesting direction where the creative and knowledge work part we thought was safe for a while. You know, I think that most people saw it as a way to eliminate repetitive tasks and create efficiency. And I feel like there was maybe an assumption that creativity and knowledge were a ways off and they're not.
They're here now. And so now, as marketers, as consultants, as entrepreneurs, we have to accept that AI is knocking at the door. It is able to do many of the things that we do. And so what remains uniquely human is probably things like empathy and personal communications and the ability to maintain and build relationships and strategy to some degree. But creativity is assisted now. It is not a uniquely human capability anymore. So I don't know. I think that's why I'm so I feel an urgency for people to understand it because you can't comprehend the impact it will have on your job and your life until you understand what it is and what it's capable of doing.
(...) Testing is critical. You have to experiment with it. And that's you know, I gave a talk a couple of weeks ago and a keynote. And I said the difference between now and six months ago or even six years ago when I first started doing talks about AI is you can leave this talk and go get tools and test it for yourself. You can get daily, you can get Jasper, you can get a test kitchen from Google like you can go experiment with it. And that changes things because now you realize the power it has and you can start connecting the dots yourself of what it means to you rather than me standing on stage telling you that.
Karine Abbou: One last very generic question. Hearing what you say, don't you think? And I'm and I'm really putting myself into the shoes of a content entrepreneur to show you the term. Joe, don't you think that instead of us not hiring a designer anymore or not hiring a content creator anymore, don't you think that it's just going to be another way for us to use them and for them to provide their job to us? Here's the point. Understanding and testing all those new things that are coming up, all the tools that you mentioned in your book, getting into the Metaverse, and also into NFT is into Web three, into all those things is extremely time-consuming at some point. So if you're spending most of your day testing all those things to figure out if it's smart for you to use them or not, don't you think at some point it might be better and helpful for people designer to use daily as something to increase their productivity in terms of producing or really as us to learn in and to do everything on our own without hiring anybody else.
Paul Roetzer: I think it'll work both ways. They're absolutely designers, illustrators, and creative professionals who are going to seize this opportunity to enhance what they're capable of doing, do things more efficiently, and expand their services and offerings. So it's not that designers in those are going away, but there are many things that you and I may be as. I know I'm not a designer that I could not have done by myself six months ago, but I can do now better than most designers I would have hired with no coding ability, no design ability, no nothing. It's all about just prompting the engine. And I would I'll tell you if I talk with these researchers and the entrepreneurs building these things. What we're seeing today is probably 6 to 12 months behind what they're doing in the research labs right now. And I've seen some of this stuff. It's the no-code ability, the ability for the non-technical person and designer to generate outputs that are indistinguishable from top designers is around the corner in many cases.
And so that's yes, we're there are still instances where we want to hire out. I don't know what those are going to be because examples like building a presentation deck, I don't need a designer and I don't need stock photos anymore. I just tell the machine what I want the image to be. Here are the words from the slide. Give me an image if I want. Posts images from my blog post. I don't need professional help for that. I don't need stock photos anymore. So like there are specific instances and use cases where the need for professional support is gone. And so I think it again is that beginning it's a moving target, but I think it's critical that creative professionals. Understand that this is coming, whether you want it to or not. As I'll say on stage, my wife was an artist. My ten-year-old daughter is an aspiring artist. I am a writer by trade. I am not an A.I. person by trade. I don't have to like this. You don't have to like it. But it's here and it's impacting our careers whether we want it to or not.
What is the impact of artificial intelligence on the creation of visual content?
Paul Roetzer: Shutterstock launched an aid division last year and acquired three companies. They there their images their stock photos were used to train daily. So Shutterstock play was actually to become a data company and use all the data they had, including copyrighted data from photographers and miners to train machines that could in theory, replace the need for them. Unsplash does not allow air generate images on their site right now. Getty doesn't allow ai-generated images on their site. So it is a that may change in three months. Imagine that. It is a it's a wild space right now. There's a lot of unknowns and a lot of people playing catch up, trying to figure out what it does to their business model, which Shutterstock moved early.
Which brands are using AI the most ?
That is a great question. So it is we run a conference on this stuff and it's very hard to find brands that are doing this exceptionally well and are willing to talk about it. But what you can do is you can look at the major spaces like retail and e-commerce are two of the biggest adopters. So you can go in and look at the brands that have a high need for personalization and that have a lot of data and try and predict human behavior and outcomes. So we always look at those industries as starters. There's but there's, there's an innovation happening within insurance which within financial services, within health care, within manufacturing, everybody's trying to kind of target the same industries. But retail and e-commerce are our major players in the space.
Karine Abbou: And who are the major and most successful vendors?
Paul Roetzer: There's thousands of them and it's growing very quickly. Yeah. So two recent ones that caught a lot of attention on the generative AI topic was Jasper just raised 125 million Series A at a $1.5 billion valuation. They used GPT three to write copy. I'm in simple terms, stability. AI is a very interesting company. They just raised 101 million. It is not a marketing company, but they enable all kinds of marketing use cases and it's an open source platform that allows people to build on top of it. There's another one called Runway ML that again, not a marketing company but as an enabling stuff. I actually just saw something this morning. Infinite Image they released which lets you not only generate an image but then say, okay, now put it background. Okay, now expand. And then you can just keep building on this image with text prompts from a more traditional sense. Adobe has always been a leader in the space. I mean, they have hundreds of features baked within their products of that. That's a major player. And then I mean, the obvious ones like Google, Microsoft was at the platform level are all building all these capabilities. So I mean, there are literally hundreds and thousands of AI vendors building point solutions.