How to Make AI Work for Work.

From American Public Media, this is Marketplace Tech.

I'm Megan McCarty Carrino.

In his new book, How to AI, Wall Street Journal tech columnist Christopher Mims offers a guide for getting the most out of the technology.

He's compiled two dozen laws of AI to shed light on the best ways to use these generative tools.

Yesterday, we talked about how individuals can improve their productivity with AI, and today we're

digging into how organizations can use or sometimes misuse it.

One I saw recently is bosses think that AI saves their workers a lot more time than workers

say it actually does.

So I think bosses are looking for every excuse to

keep headcount flat or reduce headcount. That's a big disconnect. I mean, another big disconnect

is that when you are trying to change the way that systems work, you have to do it collectively.

It's like, great, now we have to change the whole workflow for how we do this particular

job. We're all going to have to get on board, and that's going to take time. I mean, this is why

people go to business school. It's called change management. I mean, it is a dry subject, but very

necessary. One of your laws of AI that seems applicable here is that we should treat AI as if it were an assembly line robot.

What do you mean by that?

The folks who get the biggest productivity benefit from agentic AI figure out what are the really basic tasks that it can do reliably? A lot of businesses now love using

AI-based systems to record every sales call that they make and then evaluate, well, how could we

have done that better? And I have heard this really is increasing the productivity of a lot of

salespeople because, you know, this AI that's

always listening is like, well, what about this? Or other salespeople have found that this works.

...