Post

The A.I. revolution is stalling because companies don’t want to admit that integrating the technology is expensive and slow and requires human effort.

Ms. Averill is a former chief information officer of Lululemon.

I served as the chief information officer of the athletic apparel company Lululemon for eight years. Deploying new technologies, artificial intelligence included, was part of the job. But when I left, my successor was handed a new title: chief A.I. and technology officer.

I winced — not about the person but about the pattern. Companies these days are quick to stick “A.I.” in news releases, in earnings calls, in product names, in job postings and in titles at the top of their organizations. The real question is what strategy sits behind the abbreviation. A title costs nothing, but it doesn’t produce anything, either.

And that’s a problem. When companies mistake an announcement for a strategy, they spend real money — and sometimes cut real jobs — chasing a future that isn’t actually being built. And in doing so, they’re persuading Americans to reject a powerful innovation that could do so much good for all of us.

I have spent 30 years in technology, and like many of my peers, I used the newest tools to solve hard problems. In the internet era, my colleagues and I built a tool that let every customer check the inventory in each of Nordstrom’s stores. It increased sales significantly. When I was at Lululemon, we found a way to use A.I. to help executives better predict which products would sell in which stores. It took time, lots of people and big budgets.

Then ChatGPT arrived. This new technology was easy to understand for most nontechnologists — a chief executive, board member or investor — a healthy portion of whom began demanding their organizations immediately embed the technology throughout their operations. A.I. companies started popping up like food trucks at a street fair.