(Photo by Ginger Jordan on Unsplash)

Per a recent article, Johnson & Johnson has dramatically trimmed its list of AI projects:

The “thousand flowers” approach involved a number of use case ideas germinating from across the company, which made their way through a centralized governance board. At one point, employees were pursuing nearly 900 individual use cases, many that were redundant or simply didn’t work, he said. And as the company tracked the broad value of AI, including generative AI, data science and intelligent automation, it found that only 10% to 15% of use cases were driving about 80% of the value, he added.

(Source: WS, "Johnson & Johnson Pivots Its AI Strategy")

My take? This is a great move. And it sets an example for every company getting into AI.

Here's my breakdown of the two key lessons, plus my thoughts for moving forward – one concern, and one warning. You'll want to keep these in mind for your own projects:

Key lesson 1: Every AI project is an experiment. There's no guarantee it will work out.

One key to AI success, then, is to constantly review your ongoing projects – at every step you decide whether to keep going, change course, or shut it down. You don't want to pour more resources into a project that's not working out.

It looks like J&J has taken that approach. According to the WSJ article, they determined that a small number of projects accounted for most of the value – and they have culled accordingly.

Key lesson 2: You need to track metrics in order to triage. J&J was able to calculate the value of their AI work, which guided their decisions on which projects to move to the next stage and which projects to halt. You'll need to do the same.

The concern: That original pool of 900 use cases sounds like a lot. I'm sure some were great ideas that simply didn't pan out. That happens. But I also ask how many of the redundant or failed ideas could have been caught before formally launching them into pilot projects.

The warning: Keep a close eye on the winners. Sometimes an AI project only looks like it's working. The more a project seems like it's destined for victory, the more scrutiny it deserves. You'll want to rigorously test it to confirm your belief.