Confront the fear and embrace the power of control of AI
What’s the fear?
Since its release in November of last year, ChatGPT has taken the world by storm. Now in its 4th iteration, a mere 5 months later, the buzz in only growing, as people talk about its uses in healthcare, science, research, education, marketing, advertising and PR, among other things. ChatGPT has launched a vigorous discussion about Generative AI – AI that can produce things – and its role in our lives.
We know from Pew Center research that people fear it in healthcare, yet also believe it can help in surgery and pain management and much, much more. They believe it will play a critical role in discoveries and education. But many are wary despite the fact that AI already is in our lives with things like Siri and Alexa.
Quizzed on the role of ChatGPT in transforming journalism, only 16 percent of Pew respondents see it as a major advance. More than 25 percent see it as a minor advance, with the rest somewhere in a squishy middle. I raise this because these attitudes probably reflect the same kind of uncertainty among marketers. Some believe ChatGPT and Generative AI can fundamentally replace us. Others probably see it as no big deal and a majority are somewhere between.
And through it all, we are seeing discussions on things like the inherent bias of machine learning, the ethics of the platform, how we will govern its growth and other questions – some serious, some not – on its implications on our lives. In fact, as I write this, I am listening to a Squawk Box discussion on CNBC extolling the virtues of AI to radically transform the microchip industry, fostering new hope, but also creating fear as the hosts wonder what it means for humanity.
Where does that leave our industry? Somewhere on the spectrum of inquisitive to hysterical. My first experience with fear-based predictions resulting from technology displacement was the advent of the PC, which promised to lead to a paperless society. In the intervening period, truly disruptive technologies, like the Internet, completely transformed things like retail, banking and finance, and the media industries. Blockchain is seen as a promise – a promise to give us a decentralized currency regime, seen as a panacea to some. Never mind that crypto currencies are prone to 90 percent value swings and that displacing the dollar as the global reserve currency might crater the US economy, rendering our life savings worthless. Because these things aren’t perceived as an immediate threat to our livelihoods, we largely cheer them on.
But Generative AI presents a real and present danger in the thinking of some people in marketing. And pundits like Jim Cramer contribute to this when they claim that it will “level” – as in wipe out – advertising employment, while extolling the virtues for companies like Nvidia, which make the underlying chip technology. This isn’t to argue that Generative AI might eventually replace us and what we do. But is it going to happen soon? Unlikely. While it is impressive that Generative AI can access and organize ideas into coherent output, on closer inspection these outputs lack the heart, soul and overall value of human produced thinking, content, collateral, and written assets.
Where is the opportunity for marketers?
Rather than focusing on the potential of being replaced by Generative AI, we should instead be focused on Performance AI – the kind of science that allows us to access and organize massive amounts of data into actionable insights to underpin better marketing strategies and execution. Performance AI is the near-term future – one we should embrace and nurture to our collective advantage.
Marketers need to avoid getting wowed by what the technology can do and instead focus on how it can help their business. For example, it’s neat that a piece of AI can write a press release, but PR practitioners are far too focused on whether that will replace them, as evidenced by the myriad coverage of it. Same thing for creatives thinking they are going to be displaced. Or clients who are thinking of these capabilities in terms of efficiency rather than business outcome. So, the number one thing is, don’t get caught in the noise.
Instead, marketers need to be evaluating and using AI from the perspective of how it informs business strategy and, by extension, marketing and communications engagement. They need to identify platforms and technology solutions that can give them a view of their key constituents and conversations relative to their businesses and business goals. Preferably solutions that are dynamically processing information across wide geographies, topics and constituent groups. They need solutions that can easily collate and organize data so that it can be molded into actionable insights – that is, executive summary-type analyses with accompanying points of views and action plans rooted in the data. Finally, they need flexible solutions with adaptable dashboards that allow the insights to be customized to specific and changing requirements.
Based on these insights, organizations are more empowered to develop the right mix of marketing solutions and a strategic ability to address key constituent groups via the right strategic mix and channel. The result is a far clearer path to achieving a 10x multiple in impressions at a reasonable investment level.
Ultimately, these are the key differences between Generative AI and Performance AI. The first focuses on outputs and is being leveraged as a way to cut manpower and cost (at least for the moment). The second focuses on a holistic view of business, marketing and communications strategy for agile implementation and lasting impact (with a focus on augmenting growth as opposed to replacing human capital to cut cost).
Generative AI, should it reach its promise, has the potential to be a disruptive technology as defined by the late Harvard Professor Clay Christensen (as he articulates in this 2015 HBR article ). But Performance AI, while not necessarily disruptive by that definition, already has the potential to radically transform the way that we do business with clients and, done right, not displace people in our profession, but instead enhance their performance.
Where do we go from here?
My past year has been spent with colleagues assessing, evaluating, and experimenting with a wide range of solutions that use natural language and machine learning to access an unimaginably large amount of data on key topics relevant to our clients and organize it into bite size, actionable data sets. Based on this, we have developed a point of view and a solution set that harnesses the power of AI to make us smarter and deliver better results. We have coined the phrase Performance AI because we see it as AI that can drive better overall performance.
What we have learned, in this process, is that nearly everyone is attaching the moniker AI to their technology – just as everyone applied the term Web 3 to technology last year and the year before. Just because it is called AI doesn’t mean it is AI. Sometimes a social listening tool is just that. That’s not to impugn the value of social listening tools. But merely to point out that for AI to be AI it needs to have certain technology-enabled characteristics that allow it to process with logic and scale an output.
What we have also learned is that there are certain redundancies and gaps within the technologies specific to our industry. In our view, no single solution can deliver a holistic solution. We started with wanting to answer the following question: What kind of insights do clients need to take measurable action to drive marketing programs that support real business results? Through this lens, we were able to identify a suite of solutions that could be fashioned into a high-performance dashboard with outputs that answer key questions across constituent groups ranging from investors to customers, to employees, to thought leaders. Every consulting group needs to find its own right solution specific to the client problem it is trying to solve. We recommend starting with the question.
Another key learning is focused on build versus buy. The first impulse it to develop a homegrown AI solution. We curbed that impulse early, recognizing that technologists are better at technology. We should focus on what we do well – marketing – and leverage solutions from a range of partners to develop best-of-breed solutions specific to making us better marketing partners for clients.
We are in an invigorating time, when the potential for AI to revolutionize our lives is great. As marketing professionals, we have the opportunity to approach this in fear or with an attitude of discovery and innovation. My preference is the latter. Will there be unanticipated obstacles and pitfalls along the way? For sure. But the ability to give our clients better service while engaged in great discoveries is too much to pass up.