There Is No AI Strategy. AI doesn’t solve anything but is an opportunity to solve many things.
I’m noticing a weird tension building as marketing teams across all manners of business experiment with Generative AI tools. It boils down to this: Generative AI solves problems you don’t know if you even have.
Last week, a CMO at a large tech company told me they felt woefully behind and struggled to understand where and how to integrate generative AI at the marketing team level. “We just don’t get it,” they said. “Corporate is still afraid of losing our secrets to the public-learning models, but nobody wants to invest the effort into building our own. So, meanwhile, we’re just playing around.”
That hurry-up-and-wait pressure is common. In some cases, immense pressure develops to create what might be called an “AI strategy.” I’ve seen many business leaders in startups, mid-sized businesses, and enterprise companies scramble to explain their plans to their major stakeholders.
One large nonprofit threw a customer service chatbot on the website so they could brag to the board that they’re “doing AI” while they quietly debated what AI really means to their comunications strategy. Another technology company seeks a “chief of AI” to signal investors they take integration seriously.
Organizations generally think they should have some miraculous new capability. The promise (or warning as the case may be) is that AI will take jobs, expand creativity, and inspire companies to realize they don’t need those pesky humans running around doing things. And so, marketing leaders hear, “Tell us which one of those things it will be. And do it quickly! Before we’re left behind.”
Gen AI is the new car you didn’t ask for
Now, it’s not that you can’t do interesting things with the technology. Yes, Generative AI helps you express ideas more quickly or thoroughly. You can “talk” to documents, automate communication workflows, translate, summarize, and structure data. In other words, generative AI takes your ideas and expresses them exponentially faster and at scale. But there is no AI strategy. And there can’t be really.
We’ve all been under increasing pressure to do more and more over the last decade. Since the first Content Marketing World in 2011, I’ve heard marketers clamor for the proverbial “faster-horse” technologies. The fact that “more content” vs. “quality content” is still a debate demonstrates this. But with AI, we all got a new car that we weren’t really asking for, and it’s a rental.
Is it any wonder businesses aren’t sure how to feel about this car? Sure, it improves with every driver who takes a turn at the wheel. But you also have real concerns about the implications of driving this public vehicle. Does sharing your information run afoul of legal, regulatory, or competitive concerns? Also, a communal car prevents you from differentiating and building trust with your audience. So, should only a select few get to drive it?
OK, so you’ll build your own car – and integrate it into the fabric of your business. But wait a minute. If (and it’s a big if) you have enough training data to build a custom AI learning model, it may take months and possibly millions of dollars to do it right. And if you just use your “small” data set, the answers aren’t nearly as cool and powerful as something like ChatGPT.
All those considerations leave most businesses simply pawing at generative AI like a cat poking a ball to see if anything interesting comes out the other end.
What should you do?
Innovation vs. invention
While the innovation of generative AI is a breakthrough, the true functional and valuable AI-powered inventions are a work in progress.
Generative AI is a true innovation. It improves an existing idea or product, making it more efficient, effective, or accessible. Invention, on the other hand, manifests an idea or object to create something that has never existed. Innovation and invention are two different things.
In the last 25 years of the digital era, inventions that emerged from original, innovative approaches filled the world. However, many of these inventions had no link to value.
Motorola’s Iridium phone from 1998 is a great example. At that time, around 300 million people used cell phones. Motorola launched the first satellite phone to let people call from any global location. It worked fine as long as you were on a boat or in the middle of a desert. But step into a boardroom in the middle of Manhattan, and you had a $3,000 brick in your hand. The Iridium was truly an amazing invention based on an awesome innovation, but there was little understanding – and connection – to the actual value it might bring. Thus it failed.
Put simply: Innovation is not strategy. It’s a way to break things, fumble about, find new approaches. Innovation is opportunity, not direction.
What does invention vs. innovation have to do with how you get to a better plan for generative AI? Well, in order apply the innovation of generative AI and connect it to strategic value, one must fully understand the opportunities — or possibilities — of all the approaches it can innovate.
Thus, there is no AI strategy. You cannot make generative AI a thing. Generative AI is an opportunity looking for a strategy. The problem is that we have so little understanding of the approaches that GenAI could innovate.
Which approach should AI innovate?
It doesn’t matter if you hire a chief AI officer or have individuals play with the opportunities the rest of the year. If you don’t apply AI through an innovation lens, making any decisions about moving forward with an integrated approach will be hard.
For example, I recently heard from a client who wondered if they should rely on Microsoft Copilot’s suite-like platform embedded into their team’s tools or deploy a more purposely siloed best-of-breed solution for brand consistency, translation, content creation, workflow automation, etc.
My answer required two more questions. What process did they want to innovate and make better? Or more precisely, what process/approach did they want to create that would be optimized with Gen AI? And, more importantly, did they understand the current approach well enough to know where innovation might be valuable?
When asking the latter question about content creation, channel management, personalization, A/B testing, persona research, or myriad other approaches where generative AI could be a game changer, the answer was (as it often is) “we don’t know.” Thus, we just “play around” with the innovation seeing all these opportunities that don’t really look like opportunities as much as they look like that shiny new thing that seems neat.
Generative AI innovation seeks content strategy
There is no AI Strategy, just as a content and marketing team, you would never think about creating a “telephone” or “computing” strategy. Look at AI similarly. When you understand and optimize content creation, management, and measurement approaches, you can identify the opportunities to innovate them. Put simply: you don’t need a GenAI strategy. You need a content strategy that may or may not be optimized by GenAI.
To be clear it’s not about YOU understanding the opportunities. You, the person, might understand them and be screaming from the rooftops. But, the business, the institution, must understand them. It’s only then can the business understand the prioritized uses beyond how an individual benefits from using a generative AI tool. It’s only then can you can you know what opportunity creates the most value for the team, the division, the region, and, ultimately, the entire business.
Show me a business that understands that and has a shared, smart content strategy, and I’ll show you a company primed or already enjoying the innovation that GenAI can bring.. They invent a new approach, innovated with AI.
It’s your story. Tell it well.
