(Photo by Megan Lee on Unsplash)

It's said that predictions are a fool's game. That doesn't stop people from trying to predict what's next for (The Field We Currently Call) AI. I know this because they often ask me for my take.

I've thought about this question a lot. I even wrote about it in Structural Evolutions in Data, charting the many expressions of the field's overarching theme – "analyzing data for fun and profit." While it's made for some fun thought experiments, the lesson I've learned is to not make too many plans around it. There are better ways to prepare for AI's future.

Read on to understand why, and to learn how I answer "what's next for AI?" these days.

The long road is hazy

Such predictions are too easy to get wrong over the longer horizon. Case in point: in the Structural Evolutions article I posited that simulation would be next out of the AI starting gate. It was not. (It might still be someday! But if you were placing bets based on that idea, and pouring resources into building a business, you would not be happy right now.) Generative AI took on a life of its own, surprising many industry professionals in the process. Myself included.

How does this kind of unexpected change happen? Keep in mind that any number of social or economic forces can upset that "sure thing." Add someone's weekend hack project and an entire industry can crop up, seemingly out of nowhere. It doesn't matter if the dethroned champion was the logical choice; logic only gets a small vote in deciding the future.

Since the future can change direction at any moment, it doesn't serve you well to try to prepare too much in advance. Long-range plans often remain just that: plans.

Peering around corners

Plenty of companies were caught off-guard by the rise of e-commerce, cloud, or the various waves of the data field. They're understandably nervous about getting left behind – or worse yet, steamrolled – by the next tech advance. What can they do?

The solution is to be self-aware, be flexible, and stay alert. That's how you make the most of the changes on the road ahead.

• Be self-aware: Know who you are – what makes your business tick and what situations could introduce friction. (Or worse, what can take you out.)
• Stay flexible: Accept that change is inevitable, and make your business amenable to change.
• Stay alert: Keep an eye out for new developments that are already happening and growing. And prepare to move when you spot one that could impact your business.

Very few businesses saw the power of large-scale, LLM-powered generative AI at first. But imagine the executives who got one look at Midjourney or Dall-E and wondered: "what if I had access to a magic image-generation machine?" Now imagine those same executives months later, when ChatGPT came along. "What if I could generate natural text with a few keystrokes? What would that do for my business?"

You can do the same. This can be as lightweight as staying up to date on the news, or as formal as establishing a special team to keep an eye on the (near-)future. And if you have a formal risk management department, they're a good place to start. They're probably doing this already, just on the side.

You can try some scenario exercises for good measure. For example: "if the cost of [some tool we rely on] suddenly plummets, what do we do?" "Imagine a new technology comes along and eliminates the need for our core offering. How would we adapt? Would we fold, change direction, or find ways to embrace that technology to reduce our internal costs?"

If you are vigilant and agile, you don't have to make long-range plans. You can prepare for those shorter-term shifts that are coming just around the bend

This approach takes a lot of work but it pays off. Not only will you make the most of the next Hot New Thing™ , you'll reduce the chances of overinvesting in something that doesn't take off.