ADP Chief Talent Officer Jay Caldwell shares how even with all of AI’s strengths, there’s nothing more important than humans and their knowledge and skills. He shares an example where his experience helped him correct AI and how human decision-makers are always going to be crucial. Jay also shares with Patricia Stark how AI tools can impact connection to your teammates at work. For more on ADP’s AI and HR tools, visit ADP.com

Video Transcript:

00:00:00:02 - 00:00:20:14
Jay Caldwell
The deeper you get with AI at times, the more isolated you're getting because you're just working with a computer and you may forget to collaborate with others or ask that knowledge question to the person sitting next to you, as opposed to the AI that might know it.

00:00:20:16 - 00:00:31:20
Patricia Stark
So you've really spent your career thinking about talent and people at work. So what has really surprised you most about how AI is affecting people at work?

00:00:31:23 - 00:00:58:06
Jay Caldwell
A lot of things, I think first off, its pace and in two directions. I think what's interesting about AI is some initiatives move lightning fast gene AI initiatives as an example that use just general purpose tools, like the chat tools that are that are available to do many of the many different tasks. Those things have moved incredibly quickly to to workforces like ours.

They help with everyday things reading, summarizing, I should say, writing some analytics, things like that that hit an organization very quickly. Other things move a lot slower, and I think that's as complexity increases. And if the AI you're building requires multiple data sets, multiple pieces of information, those move a little bit slower. What surprised me more recently, especially given some of the headlines early on as AI has grown so quickly, is it's not a great tool for everything.

And some of those things are tasks that need to be very concrete, very specific. Payroll, as an example or very human oriented, where you want to make sure the connection of of people is very strong, or making people decisions, which is what we do in our space very often. It's not great at that. It can assist you, but not something you would rely on.

How about you? What surprised you?

Patricia Stark
Yeah, I think what surprised me is that AI gets it wrong every once in a while and I have to really check it. And sometimes I argue with it. Sometimes I'm like, no, I know that that's not correct. And then it will apologize. It will say, oh, you called me out on this or you corrected me.

Jay Caldwell
What's funny, though, is when you do correct that, don't know if you have this experience. It compliments you, which makes you feel awesome and truly the expert.

Patricia Stark
But there's something weird about that too, because I feel like it tries to hard to please me. And I think there could be good and bad. And I wonder if there's a like a negative to that.

Jay Caldwell
There could be. Actually, it reminds me some of the research at ADP performed found that the more people use AI, the less connected they felt to their work and teammates. So that highlights maybe some of the downside, which is also another surprise, is, you know, just intuitively that doesn't make sense immediately. But when you think about it, the deeper you get with AI at times, the more isolated you're getting because you're just working with a computer and you may forget to collaborate with others or ask that knowledge question to the person sitting next to you, as opposed to the AI that might know it.

Patricia Stark
Yeah, I think that you can't just see what AI gives you and then just accept that at face value, like it's the be all end all. You really have to say, is this correct? And is this what I asked it to do? You have to check that accuracy.

Jay Caldwell
Yeah. Yeah, absolutely.