Operationalizing AI

Foreword

In my role as an industry analyst, I’ve been studying, advising, and helping to educate organizations adopting machine learning (ML) and AI for much of the past 10 years. For half of that time, I’ve been hosting the TWIML AI podcast (originally “This Week in Machine Learning & AI”), which I launched as a way to make my learnings and work in the space more public and accessible. I’ve now had the benefit of speaking with nearly five hundred of the world’s leading experts on various aspects of ML and AI innovation and adoption.

Given this background, and my enthusiasm for the opportunities available to organizations through artificial intelligence, folks are often surprised to hear me proclaim that, in spite of all the hype, you don’t actually need to do anything differently for your company to benefit from AI. That’s right: you don’t need to hire data scientists, collect or label training data, or build any machine learning or deep learning models.

Rather, it is already the case—and it will be increasingly true in the future—that AI will power many of the products and services that your enterprise uses, independent of your own investments in the technology. In other words, your organization is a passive beneficiary of AI.

Of course, there’s a catch. These passive benefits are a rising tide that floats every boat. They are undifferentiated—your enterprise gets only what is available to everyone else in the marketplace—and as a result, they are also undifferentiating. That ...