Quartz's Colleen Cabili wrote a fascinating story the other day about companies like Ford and IBM re-hiring workers they had recently fired, ostensibly because AI could do their work faster and cheaper.

The results from implementing such plans weren't quite what the companies hoped, to say the least.

But Cabili also notes:

The reversals arrive as AI-related job cuts have been climbing. Artificial intelligence was the top stated reason for layoffs across all U.S. industries for the third consecutive month in May, with AI-related cuts reaching 38,579 — about 40% of all announced job cuts that month, the highest monthly total since outplacement firm Challenger, Gray & Christmas began tracking the category in 2023. Companies have also drawn scrutiny for using AI as a public explanation for layoffs driven by other factors, a practice some management experts call "AI-washing."

This looks like a real paradox. How can some companies be disappointed by AI exactly at the same time others are citing AI as the reason to make new rounds of layoffs?

To use a fitting but much-derided word in other contexts, I think the answer is that we're at a real transitory phase regarding AI and employment. It's a transition from mania to pragmatism, from hype to firsthand lived experience with this tech inside enterprises.

That's the narrative I see lurking between the lines of Cabili's story, and it's ultimately a very healthy one.

So does this mean AI sucks as a technology, or that it won't have big long-term effects on the economy? Not necessarily. Not at all. It just means AI is more like a cool new tool that has both legitimate utility and real limitations, just like every other tech tool that's come along in the past. Get real.

In other words, startup founder and OG blogger Anil Dash was right all along. His pragmatic take on AI back in October looks more prescient by the day.

I should also mention here that I'm eager these days to work with companies that want to implement AI and other emerging technologies in a sustainable way. No hype. Just real use cases and real growth. If that describes you, let's talk.