There’s no doubt Artificial Intelligence (AI) is one of the most transformative technologies of our time. From self-driving cars to virtual personal assistants, AI is changing the way we live and work in countless ways. While the potential benefits of AI are enormous, there are also significant challenges that come with this new technology. Here, we explore how AI is shaking up the marketing industry, its benefits and challenges, the role of humans in its development and use, and how we as an industry can prepare for the future.
AI has long been a topic of much discussion at dining tables, writing tables and board meetings. But prior to the end of 2022, for many of us, AI worth worrying about still seemed like science fiction.
Then in November 2022, ChatGPT was launched, and suddenly it became clear that the time to start taking notice was now. Questions of privacy, ‘bad actors’, dangerous biases, and ethics began to be asked as we grappled with this next stage of societal evolution.
In fact, as this article is written, OpenAI CEO Sam Altman is testifying before Congress to decide the way forward for AI regulation. No doubt by the time this goes to print there will be further developments, which is testament to just how quickly things are moving in the space right now.
Reactions to ChatGPT are often mixed. Even Sam himself has publicly declared he is “a little bit scared” of what he has created. But based on data from Finbold, AI is here to stay, with the estimated market share of AI in 2023 standing at $207.9 billion.
By 2030, projections indicate a remarkable surge of 788.64 percent, reaching $1.87 trillion. AI’s market share is expected to surpass the $1 trillion mark as early as 2028.
To delve into how our local industry views this technology and its potential use, a group of researchers from AUT’s School of Communication Studies have been working on a project around the use of AI in advertising, interviewing New Zealand agencies of varying sizes, to find out what is possible now, and how it may be used in the future.
Researcher and Senior Lecturer Justin Matthews says one of the most common concerns raised during these focus groups was that AI now has the potential to take people’s jobs.
“I think in the context of the advertising campaign, the capability [of AI] struck them as a bit of a wake-up or a moment to go, ‘Okay, things have shifted’, where the campaigns they saw in front of them, the general experience across all the focus groups was, ‘Okay, this is quite impressive that it’s able to do this kind of thing now,’” he says.
“There was definitely thinking around what this [AI] will do as this continues to expand and get more sophisticated, in terms of the way roles will work inside the creative team structures that advertising agencies use.”
With AI technologies able to create music, text and even images, that on the surface look as though they could have been created by a human, the researchers conducted a series of focus groups with New Zealand-based agencies to explore how they felt about the capability of Open AI’s DALL-E2 text-to-image technology.
To gauge the views of these advertising creative professionals, the researchers selected a campaign that could be recreated using the text-to-image engine DALL-E2, one that relied on images and one that would be familiar to advertisers. The Volkswagen newspaper campaign ‘Small but ferocious’ was chosen ,as it blended animals as a metaphor for the economical but powerful TSI engines.
In this campaign, the animals are depicted as though they are photographs, so the AUT researchers used DALL-E2 to recreate the original image of a Baboon/Hummingbird along with several other animal combinations. In about 30 minutes for each execution, DALL-E2 had recreated a visual for the campaign.
Taking these images with them, the team from AUT formed five focus groups, consisting of creative directors, art directors, designers, copywriters, content creators, strategists and account executives, and interviewed them to find out how they felt about the work produced by AI.
The research found that initially the participants were impressed with the aesthetics of the pieces but on closer inspection once they had sat with the image a little longer, they noticed a lack of a certain level of craft.
Despite not being quite up to standard, many participants recognised the potential for the AI
to enhance the current image production processes and generally speed things up.
One participant was recorded as saying: “they’re not ready to go to print just yet. But, I reckon an hour each, and a proper background, they’ll be ready to go.” Their colleague echoed this sentiment saying: “the AI is making the key visual, but then we still need our design team and our retoucher to tweak it to get it right.”
Another benefit identified by the participants was generating artistic ideas and experimentation, playing with out-of-the-box thinking, and reducing the time it might take to do these activities.
As one participant said: “[T]he main thing is just the fact that it does cut down time and you can spend more time doing other things that you might be wanting to spend time on, working through the overall concept I guess.”
Privacy when dealing with sensitive commercial information and customer data has long been a concern of the marketing world. When marketers use this publicly available technology, they are going outside of their corporate boundaries, which raises further privacy concerns.
Recognising this, particularly within the creative industry, tech companies such as Salesforce and Adobe are creating their own open language models, meaning marketers can enjoy the benefits of using AI while staying within their corporate boundaries.
Salesforce has been using AI for some time already, and it has already changed the way it operates thanks to a platform called Einstein AI, which delivers AI through all its applications across sales, service, marketing and commerce.
However, in March 2023, Salesforce launched a new programme called Einstein GPT, which is the company’s first foray into generative AI and is taking things up another level.
Generative AI is the class of artificial intelligence that allows the user to generate content such as text, video and images and can learn from the information put into it.
Rowena Westphalen, Vice President, Solution Engineering, Australia at Salesforce says this means marketers can be comfortable in the knowledge that their corporate information is staying within company firewalls. And because generative AI learns, there is the option for it to learn from an individual corporation’s data set, meaning it can be consistent in tone and highly relevant.
Adobe was also quick to recognise the power of AI with the launch of the new Adobe Sensei GenAI services in Adobe Experience Cloud. Integrated as a marketer’s co-pilot, Adobe similarly pitches this technology as boosting productivity and creative control while ensuring trusted governance.
Similarly Adobe introduced Adobe Firefly, a family of creative generative AI models integrated into Adobe Experience Cloud. This generates safe content for commercial use. Trained on professional-grade Adobe Stock images, openly licensed, and public domain content, Firefly output offers real business value without infringing on others’ intellectual property.
Sharn Piper, CEO of Attain, a sales, media and marketing agency, has been using multiple AI tools for the past 12 months or so. In early May, Attain joined forces with a company specialising in marketing technology, Robotic Marketer, allowing Attain to integrate this tech into its selection of consulting services.
This means that instead of using multiple tools to put a strategy together, all the necessary information is in one place on one platform, speeding up the research process. This in turn allows Attain staff to spend more time being consultants rather than researchers.
“We do a lot of marketing strategies and we did a lot of them last year, particularly with the businesses going through Covid interruptions, and what we found was a good strategy could take six weeks to develop,” Piper says. “By the time you’ve done the research, the analysis, you’ve put it all together and put it into a nice document that everyone can read, it does take a good amount of time. [Robotic Marketer] enables us that same quality, if not a little bit better, because of the analysis it uses, in the space of about two to two-and-a-half weeks.”
Looking ahead, Sharn believes there is no doubt that AI is set to be a major disrupter. “AI for the industry is almost like when builders got their tools.
“Something might have taken you eight weeks to prepare, now takes us two. And eventually I think it might even get less than that.”
But assuming a future scenario where using AI is the industry standard, and marketing becomes more efficient, more creative, and more strategic, how can marketers stand out from the crowd?
“It comes down to humanity and trust,” says Salesforce’s Rowena. “A lot of the conversations I’m having with our customers is how we humanise digital. The opportunity for marketers with AI is to be able to drive one-on-one personalised engagement with their customers at scale. Rapidly expand revenue, reach out to more people than they had. The downside to that is how do they preserve the culture, the brand, the unique feel that makes them them, which has probably built the trust with the consumer at the same time?”
Dr Jonnie Penn, Professor of AI Ethics and Society at the University of Cambridge, offered a more skeptical view of a future with AI, when he spoke at the Future State event held at Spark Arena in May this year.
During his keynote presentation he quipped that some might describe ChatGPT as “mansplaining as a service”, referring to the confidence of which it spouts information regardless of whether its factual or not.
Simply creating a technology that is cognitive does not equal intelligence, he points out, before quoting French cultural theorist Paul Virilio who said: “When you invent the ship, you also invent the shipwreck.”
Jonnie believes the way we should be tackling AI is similar to how music is written with notes and rests. “You have to deliberately choose to pause so when the note comes in it has some emotional resonance,” he says.
“I think the era of digital technology that we have grown up with and that I have grown up with has been maximalist… We talk about data like it’s the next oil. I think we are going to look back on this as antiquated because the environment can’t stomach it, and it’s actually not the best use of these tools. So as in music I’m advocating for ‘rest engineering’ where we start to deliberately isolate the areas that it’s just not needed.
“I think this is the most pro-AI position you can take, because if you want these tools to become part of
the cultures that we live in, we need to reduce our exposure to risk and start with the areas we can get the most value from.”
In a local context, Spark NZ is using it as a test case to show how machine learning and AI can produce real results for New Zealand businesses.
With its robust machine learning model, Spark NZ can predict who needs what service or product with 80 percent accuracy, enabling personalised targeting and significant conversion rate boosts. By using AI, Spark is creating real-time, one-to-one segments across millions of customers.
This approach also streamlines the mundane marketing tasks, freeing up humans for more rewarding, creative work, driving sales, and enhancing customer satisfaction.
Spark Marketing Director, Matt Bain, says Spark NZ was an early adopter of this technology as it was clear this was the way the world was headed.
“The reason we got into AI early, three years ago before it was really mainstream, is it was inevitable that it was going to be a thing. And if it was inevitable, then why wait?
“I always like to think that disruption is either a good thing because you’re disrupting by being first, or it’s a bad thing because someone’s disrupting you. We want New Zealand business to be on the right side of disruption, which is leaning into it responsibly and benefiting from the compounding rate of innovation.”
Of course, with this adoption, there will be a steep learning curve. Matt says that Spark NZ works with its business customers to ensure their data is clean, useable and that their privacy controls and capabilities are in place and up to scratch.
“The first thing is to get your data in order and then decide what you want to achieve. You don’t want a solution looking for a problem.”
Simon Bird, Chief Product and Strategy Officer at PHD Media, says the beauty of AI might actually lie in its stupidity – somewhat ironic for a technology designed in the pursuit of intelligence and named accordingly.
He believes AI is underutilised when coming up with creative strategy “not because AI is any good at creative strategy but because AI can be really stupid in how it puts things together”.
“You can see this in AI visuals, where AI makes mistakes that a human being never would, like creating hands with nine fingers or showing people eating unpeeled bananas via their forehead.”
While these types of mistakes might be bad in some industries, for a creative industry like advertising, “incongruence is often the secret to strategic and creative leaps”.
“With the right prompts AI will put concepts together that are so ridiculous a human would not have thought of joining them together. Many of these will remain ridiculous but some of them will be strategic or creative genius.”
Regardless of whether you are a ‘tech optimist’ or you’re more on board with Jonnie’s ‘rest engineering’ philosophy, it looks likely AI is set to make a drastic change to society. So buckle up and remember your manners when using ChatGPT, because you never know – the robots might soon be taking over.
This article was originally published in the June/July 2023 issue of NZ Marketing. Click here to subscribe.