• by J. Walker Smith , Op-Ed Contributor, 4 hours ago

The buzzword in marketing right now is AI. Well, acronym, that is. But either way, AI is all the rage.

Every firm of every sort is working hard to incorporate AI into every process, every output and every pitch. It’s AI or bust. It’s not clear if it’s tulips or not, but interest in AI is unprecedented and shows no signs of abating anytime soon.

There is no 21st Century business trend that looks like it. Some quick work with Google Trends finds that compared with AI, peak interest in Big Data, cloud computing and social media barely even register. Bitcoin holds up, though just barely. Only mobile shows anything near the interest of AI.

I asked ChatGPT about these comparative trend lines. Its takeaway is that AI is a general-purpose technology with everyday applications. All the others are specialized. Except for mobile, which, in its own way, is more like AI’s universal presence, impact and functionality.

I agree with ChatGPT. I think that’s the best way to think of AI—a general-purpose technology that will serve primarily as a platform for applications and derivative technologies. We will all use AI, but not AI per se. Rather, AI will be embedded into the things we use. 

This means three things to me.

First, the benefit of AI is making the benefits of other things better. AI is a platform not an end-use itself. Just like mobile. The value AI brings is to make shopping easier, so the connection with consumers is easier shopping, not AI shopping.

AI won’t change the benefits that people seek from products and categories. It may change which brands deliver those benefits better and it may change the quality of experience for a brand or category. But there’s no new benefit that AI will bring to a category.

This is the soft underbelly of AI. It may not take AI to deliver something better. The leisurely family stroll through the Saturday morning farmers’ market that is part of the pricey neighborhood vibe may be beyond anything extra AI could deliver.

In other cases, something other than AI may be a better way of delivering the relevant benefit. There’s nothing AI can do about legroom on a plane or the original formula for a soda. Experience and taste will often trump anything AI can deliver. AI will not always be the differentiating factor for consumers.

Second, as a platform, AI will be available to every company. Some will use AI better, of course. But advantage won’t come from AI itself. It will come from the things that generate and facilitate better usage of AI, like innovation, leadership and organization.

However, best practices travel fast. Companies learn from one another. Key executives switch jobs and take skills with them. Conference presentations show off the successes of one company to every other company in attendance. Consultants learn from working with pioneering clients and share that learning with all clients.

As companies get up to speed on AI best practices, AI will become table stakes not a source of differentiation. Given the take-up rate of AI and the ongoing experimentation with applications, parity of competencies is likely to happen faster with AI than with past technologies.

Companies that fall behind will suffer competitively. But companies that keep up will be the same as every other competitor. A few companies always stand out with novel applications, but nowadays, and especially in the fast-moving marketplace of AI functionality, these gaps close quickly.

More than anything else, AI is likely to deliver a more advanced marketplace in which stubborn inefficiencies have finally been fixed for every brand and product. It is unclear if there is any potential for AI to enable one company to assume a commanding position when all its competitors are AI-proficient as well. The AI future may simply be a mirror image of today’s market share and profitability, just with AI instead of without.

I call this the paradox of quality, which is nothing new. Every product and service today is better than the past. Quality is greater across the board—because quality is contagious. Once one company finds the best way to do something, all companies follow suit. The result is both higher quality and greater parity. That’s the paradox—over the long-term, differentiating innovation works against sustained differentiation. The thing that really changes is that the cost of doing business gets bigger.

Finally, AI is hard to layer onto something already in place. Which is a challenge for big, established brands. Big brands will be the eventual winners in the AI race, but they won’t be first out of the gate with the biggest breakthroughs. Because big brands can’t afford to experiment with platform shifts.

Over the immediate term, AI-native firms will get more out of AI and do more with AI, as big brands watch and learn to later acquire and integrate. Big brands can afford to be behind at the get-go.

A new paper from a pair of researchers at INSEAD and the Harvard B-School finds that advantages of AI mostly accrue to AI-native firms not established non-AI startups. AI-native firms employ 25% fewer people, have 13% more engineers, and have 15% fewer entry-level and manager-level employees. Making AI-native firms flatter by half a seniority level. AI at AI-native firms is also more likely to be embedded in both process and product, not merely layered on top of existing workflows.

This isn’t anything new to AI. We’ve lived through past hockey-sticks of digital-native, social media-native and internet-native firms. We have seen that there is a good bit of truth in the idea that native firms have more savvy with new technologies.

But as a colleague of mine once observed, there is a difference between managing for value and managing for valuation. Native firms tend to be the latter, looking for a payday that transfers their systems, know-how and people to a bigger home.

Out-of-the-box AI innovation is unlikely to come from big, established brands. They have built their processes and products on existing general-purpose technologies. It’s impossible to justify breaking that mold for something untried and untested. AI-native firms will sort out what’s what, then big brands will upgrade and adapt.

We see a lot of AI action these days with firms big and small. But the future is a platform shift like the move to mobile, not an operating update. Platforms are fully embedded in the operating structures of big brands. Big companies are not going to leap out of the box just yet.

AI is changing so rapidly that big moves for big brands are risky. For example, brands are busily adapting to stay visible with the text-based LLMs that are powering consumer search and recommendations. But such LLMs are certainly not the best that AI can do. There will soon be AI technologies better than text-based LLMs. Going all-in on LLMs as they exist today is a bet that the AI future has arrived, when in fact the future has barely begun to unfold.