Katherine Jones - The Hidden Costs of AI in HR

[00:00:00] Welcome to the HR Data Labs Podcast, now part of the WorkDefined Podcast Network. Join us as we explore the vital role of compensation, strategy, data, and people analytics in navigating today's complex business world. With the resources of WorkDefined, we're bringing you deeper insights and actionable ideas from top HR experts. Now, here are your hosts, David Turetsky and Dwight Brown. Hello and welcome to the HR Data Labs Podcast.

[00:00:27] I'm your host, David Turetsky, alongside my best friend, co-host, partner in crime from Salary.com, Dwight Brown. Dwight Brown, how are you? I'm good. And I think that partner in crime, given the fact that we think that I look like I'm posing for a mugshot, we're getting closer and closer on that description. Yes, we are. But we won't disclose anything that gets you closer to getting a mugshot. Yeah. Some things are getting better left on sad, right? That is true. Discretion, the better part of valor, they say.

[00:00:57] All right. Today, we have with us a very special guest, someone who I thought we'd had on the program before, but for some reason, it's escaped us, Catherine. Catherine Jones, the wonderful, the amazing, the effervescent Catherine Jones. How are you, Catherine? I have never been better. Thank you very much. That's awesome. Well, Catherine, give everybody a little bit of background as to who you are and how'd you get to this moment? Oh, my goodness.

[00:01:25] Well, I think most of you and most of the business world out there knows me as a high-tech market analyst, which I have been for since Moses was a pup, maybe. But you consulted with the pharaohs is what you're saying. Absolutely. You know, that pyramid building was really difficult. It was a pyramid scheme. It was a lot of wiring that thing was terrible. Oh, you know, Wi-Fi out there. You got about an HR issue.

[00:01:55] Yeah, exactly. But before that, I was in high-tech. And that gave me a really, really good background as an analyst, because a lot of analysts grew up and they were analysts. Right.

[00:02:09] And I have been responsible for product development and the whole life cycle and working with customers on very complex SAP projects and working with sales on how do you position things, marketing, and how do you talk about data communications when nobody understands it. Right. And working with the federal government and products that they need. And why is there a difference between Ethernet and 802.3?

[00:02:37] And in building that in the products, there is a difference. And if you're in building DDN applications, you need to know what that difference is. Wow. In addition to that, which was a wonderful background for being an analyst, before that I was an academic dean. Wow. Wow. So, a lot of people, when I was, how, oh, academia is so wonderful. Why would you leave that for high-tech? My answer to that is they're a lot alike. Mm-hmm.

[00:03:05] They're both very political. They're both very competitive. And all in all, I would say, I like people better than things, but the things that high-tech was building were really interesting too. Before that, I got a PhD at Cornell. Cool.

[00:03:27] I was a rarity there too, because Cornell has a wonderful capability of allowing you to build your own program at the doctoral level. And I really, really, really, really wanted to be an English professor. I wanted to write my dissertation. I wanted to write the definitive biography of Sylvia Plath, who, as you remember, stuck her head in an oven. Oh, my gosh. And actually... We do that every day.

[00:03:57] Yeah. Part of our job description. I know, but you know, your... I'm sorry, your poetry is just not up to Sylvia's. But I was also very practical. I had two little kids, and I was dragging around through grad school. Wow. And so I was in the School of Education. My TA-ship was in the English department. I ran their, what was called a Master's of Arts in Teaching program.

[00:04:25] So throughout my academic work, which is both in English and in things like administration and sociology of education and stuff like that, I was able to teach how to teach English to high school kids. Wow. And the students I had were all master's candidates, and they were absolutely wonderful. In addition, I got to teach English at Cornell.

[00:04:50] I got to teach writing, writing from experience and freshman comp. And, you know, we figured out that I might have been around doing that when Josh Bursom started Cornell. Wow. And freshman comp was required, but he wasn't in my class, or I'm sure I'd remember him. Wow. But I ended up with that with writing, being a good writer, and understanding academic or understanding research. Mm-hmm.

[00:05:16] So those were two skills that I left graduate school with that really served me well forever. Right. Smith be an analyst, yeah. You know, so all of that kind of built together to make who I am today, I think. Wow. Well, that's amazing background, but we still need to know. Catherine, what's one fun thing that no one knows about you? Oh, well, there is something. There's a couple somethings.

[00:05:45] Probably the most interesting one was something my parents forbade me to do, right? We'll start with that. I had a horse when I was, I don't know, kid. And my horse had a colt, which was a young stallion. And I was forbidden to ride him until he went off to be trained by somebody who knew how to train him. However impatient that I was, I broke him myself. Wow. And he was, he was a big boy.

[00:06:14] He was, he was, even as a baby, he was a very, very, very big horse, a very strong horse, a very stubborn horse. I had gone off, over his head, off his behind, every part of his body at one point or another. Wow. But he still was my horse. And at the end, I had a broken horse, a broken being a tame horse. Oh, okay. Yeah. There was one more gotcha in there.

[00:06:43] I'm glad you clarified that because we were wondering. Yeah, exactly. No. And I was going to sell him to a family that had three little kids. So they brought over the kids and I put all three kids in the back. I walked my horse around and he was just as sweet as he could be. And off he went with them. And I was crying because of course, there went my colt. Right. And he brought him back the next day.

[00:07:09] Because apparently, the one thing none of us had ever figured out was that that horse would only allow people on his back if I was there. Oh my God. Oh, really? So he was a partially broken horse. No, no, he was really broken. He was. Yeah, exactly. You know, not that everybody, if everybody in business had been as malleable as my horse was. Oh, I think they are actually. We'll have to get into that.

[00:07:38] Wow, that's cool. That's really cool. I never had a horse, but my kids have ridden me as a horse. Yeah, they broke me. Yeah. So, Catherine, we're going to have a wonderful discussion today about how to not talk about artificial intelligence in human resources. Because no matter what we talk about, it always comes back around.

[00:08:36] To AI. When we talk about the world of business today, you can't talk about publishing. You can't talk about the stock market. You can't talk about manufacturing.

[00:09:02] There's almost no industry that you can talk about without talking about the impact of artificial intelligence or robotics or both. Absolutely. My hip joint replacement is done by a robot. Really? I do not know if he has a name, but I am going to inquire. So I address him appropriately. That would be a really good introduction. Hey, who's operating on me? Oh, it's that thing over there. What's his name? Yeah, I want to know.

[00:09:33] We name our pets. We should name our robots. Especially the ones that are going to operate on you. Yes. But I mean, that's exactly the question is, have we come to the stage in our evolution? Now, with computers, where there is literally nothing that has not been touched by the world of automation. I can't think of a single thing, in all honesty.

[00:10:03] And look how long we've been talking about artificial intelligence. I've been writing about it for the last at least four years, if not five. And so it's this, it's not like a ta-da, exactly. But we do have ta-das. Right. And the change between general, gen AI and what we're seeing now is a ta-da, I think. And I think that's the thing that's really going to change a lot of, certainly a lot of the product development we see. You can't escape it. No.

[00:10:33] Anywhere you can't. I have an Alexa beside me and now she's going to say something probably. No, I turned her off. Exactly. You know, what do we think she is and how long have we all had Alexas and things like that? Well, it's at least been a dozen years. Yeah. What do we think that is? It certainly is. There's not a little person in a little, you know, that there's nobody in there talking to you.

[00:10:58] Well, and AI is, I mean, we've been, we've had AI for years and years and years. It's just, this is the, this is one of the first times where the label of AI has, has really kind of stuck with everything that goes into the, into the, the AI machine, if you will. Right.

[00:11:18] And it's, you know, the, the fact that we've had that, but people are thinking of it so differently and the technology is interacting so differently. And that seems to be sort of the tipping point with everything. And we look at agentic AI where the AI talks to AI and in a way, way back when they used to, there used to be a way of looking at software that was almost the same thing.

[00:11:45] Or there were little things that had like a noun and a verb and they would talk to other nouns and verbs. And that's exactly kind of what this is. Yeah. If exactly and kind of can go in the same sentence. One of these grammatical things I will have to investigate. Right. But, you know, it's, it's not new, but it certainly has evolved. And I think that some places like, is it sentience? I think that's a dumb question in a way. Well, I don't think that's what we need to worry about.

[00:12:13] We need to worry about it maybe, but in business, whether your AI has feelings or not, I think is irrelevant. What we need to do is, can we do a job better with tools? So a whole bunch of years ago, how, oh, I bet it was five years ago. I did a study that looked at what HR people thought about AI. And it was a very, they were very lukewarm, but they all thought of it as office automation. Mm-hmm. It will make my job easier.

[00:12:43] Yeah. In the same way that we talked about office automation, what, 20 years ago? Yeah. Yeah. They see that as an extension of the same thing. And maybe that's a very logical way to do it. But I think that it's more than that. Yeah. You know, all in all. And I kind of like it actually. Well, here's my thinking on this. And I'm sure because Dwight and I have talked about this a lot.

[00:13:09] And we talked about this a lot with Jordan Morrow as well on the HR Data Labs Brown Bag Lunch. We've talked about the concept of the AI automations helping HR today by taking some of the tasks away and by being a tool in the utility belt of every HR superhero, right? Everybody who's doing HR today has to do more with less.

[00:13:37] And so they could use artificial intelligence and whether it's ChatGPT or whether it's doing some kind of automation as a way of being able to do more. Absolutely. And I guess that is office automation. Sure. But not office automation. I think it's really job automation. It's really, really processes. And early on, I think that you could look at a lot of it as work force, work, workflow on steroids. Exactly.

[00:14:07] That's not a bad thing. No, no. But people are resistant to it, Catherine. True. Well, it's scary stuff, right? Well, we don't understand it. It's kind of like ghosts. We don't understand them. And so we're scared of ghosts. I'm not scared of ghosts. It's true. I ain't afraid of no ghost. I am. Well, I think the, I think that HR has fears kind of justifiable in a lot of ways.

[00:14:32] It makes the job easier, but the easier it gets, the fewer people you need. I mean, people now are looking at, well, can you just merge HR with IT? Because now it's a, it's just a piece of software. You don't need all these people. And that's a different, you know, I don't think it's- Is that really true though? I mean, is that really true? Right. Does AI help you get there yet? Well, if you even look at the talent acquisition programs, and that was something I started looking

[00:15:01] at when they, basically when they first came out. Sure. And they, over the years, do more and more, and then they get them together. Now I have always been an integration bigot. I will say that offhand. So instead of lots of little pieces doing stuff all over the place, I like them all together in a package. Sure. And that's what we almost have exclusively today. A startup may start as a standalone, but if it's a good standalone, somebody is going to buy it and integrate it with the rest of their package.

[00:15:29] So pretty soon you'd have a total solution kind of thing generally. And I say generally, because of course people buy point solutions all the time to do, you know, curl their hair or God knows what. Yeah. Right. But yeah, it's just, once it's together, more and more of that is being done. The calendaring. Remember way back when, when scheduling somebody that you're going to hire was so painful. Oh yeah. You wanted somebody to come in working with their schedule.

[00:15:58] You wanted them to talk to five different people. Why? You know, it was really difficult. And now that can all be done basically automatically. Mm-hmm. We started that a years back when, you know, standalone scheduling products that would look at your calendar. Oh my gosh. You can see my calendar. And I'll be cared. So I think a lot of us also along the way gave up a lot of things that we thought were very private to us. Right.

[00:16:23] As we looked at having everything connected for much easier ways of getting work done. But that's data integration, right? That's, that's being able to get authenticated on enterprise or on Office 365 or Enterprise 365 or whatever it's called these days. And yeah, I mean, there are a lot of point solutions to do that. And there are some companies like Workday who say, we want to have one vertically integrated stack. Mm-hmm.

[00:16:51] And there are others, I'm not going to name the names, but you know them, where they just buy every company and then they have to spend millions of dollars additionally integrating them so that their databases do talk to each other or they don't, but it's all behind the scenes. Right. So, but, but I mean, that's, that's still the trick and getting, getting the automation to work together. That was a long time ago. But I mean, we've been talking about that for decades in HR, you know, getting stacks

[00:17:20] to be friendly and APIs. But remember one of the reasons that people, a company buys another company isn't necessarily that I want to sell their software and work it in integration because they usually get rid of it. They either want the technologists the company has and hope to keep them or they want to buy the customer base and just get rid of that company. Right. So there's a lot of, a lot of other stuff in the background besides I'll buy this because I like it or it might be a short term thing. Sure.

[00:17:49] We'll buy Taleo and use it until I've got a better one. You know, I mean that, that happens all the time. Oh, it does. Right. Yeah. And you wrangle whatever value you can get out of the people, the customer base, the code, the IP, the patents. And then, you know, whatever, whatever fits your business model stays and whatever doesn't. Hmm. Okay. Well, we at least got rid of some competition.

[00:18:13] And then we gracefully transition all those customers into our lovely solution that they'll complain about because it doesn't work this way the old one did. Yeah. Can't we have all the functionality that was in the other one? Right. Right. Yeah. If we always did that, if we always kept everything that everybody had, we'd still have all the software from the sixties. Well, good point. Good point. And it actually worked. You can rely on it. I mean, sometimes better than what we currently have. Exactly.

[00:18:43] I mean, there are still many companies working on mainframes and you know, the mainframes still work. Right. There are a lot of them. Hitachi still sells mainframes, I think. Um, there are a lot of companies that still sell and service mainframes because it's very lucrative. Well, that's true. Right. But getting to be a precious commodity. Oh, it is. It is. But, but I mean, that's, and that's one of the problems here is sometimes we're trying to solve the same problem we've been solving for decades.

[00:19:09] It's just faster or cheaper, or it has some kind of bell or whistle. The thing about this artificial intelligence thing is, I mean, and we see this all the time in the news, especially because it's great click bait. We're going to replace humans with computers. AI is going to take your job. Come on now. I will say there's some jobs that may be eased out over time, but there's so many more jobs because this is not easy stuff.

[00:19:39] This is hard stuff. Who's going to somebody in your company. And I said that talking to, um, HR folks in a conference years ago. I said, somebody needs to know if this product goes rogue. So somebody has to understand enough of a algorithm to know what the thing was supposed to do when you bought it and test it to see if it does it. Oh, sure. Remember the Amazon thing where they hired only men, um, because that was who was successful in the past.

[00:20:08] That's a, that example still can happen or we call it hallucinating. I like the idea of the thing going rogue, but you know, it's the same concept of somehow giving you incorrect data or incorrect, uh, ideas or results. And somebody has to be able to recognize that. Right. And somebody has to know what it, cause we're talking software. It's not, we're not talking, um, people, people, right.

[00:20:37] My horse was smarter. Yeah. But I mean, and that's, and to get back to your example, the example of you broke your horse for you, well, you didn't know that the horse wasn't going to work for those kids. No, you thought it would. I did. In the same way, the AI algorithm that you've created will work one time. And then the next time it hallucinates and puts in rogue data as fact.

[00:21:07] And you go, well, wait, that doesn't compute. That doesn't make sense. Why didn't it do exactly what it did for me? Right. And the answer is because it didn't know any better. It went rogue. My horse went rogue on me. Yep. And the same thing with technology. So it's a good example there, isn't it? It is a perfect example. Yeah. I mean, I, and, and that's the piece people, people don't realize. And I think a lot of that is just because there's, you know, a lot of people don't have the technical

[00:21:35] know-how and they don't realize everything that goes on behind the scenes when you're dealing with the software system, when you're dealing with the servers to host these things, especially if companies are going to be implementing AI on internal LLMs. Well, you still have to have hardware and the infrastructure. And you have to, to your point, have the oversight of what's going on and what's being spit out.

[00:22:02] And you have to make sure that you are actually feeding documentation into the learning algorithms. Those are the pieces where we're going to see an uptick in jobs there. We're going to, we're going to see a reshuffling of the workforce. We're not going to see a replacement of the workforce. It's just reshuffling the, the chess pieces with it. Which does make upskilling when we talk about that. Yes.

[00:22:28] It's another tired word, uh, or reshilling something that we definitely need to think about. Definitely. Because there are new skills. And I remember years ago, I think it would, remember we were told, um, uh, HR people, they had to be more statistic and we were going to do analytics. And they said, we are people, people, we don't do math or whatever. And it's so scary. Right. And I do analytics and the analytics that they looked at were things like days to hire. Yep.

[00:22:55] This is not rocket science analytics. Right. This is something that the average person can very easily digest and a little chart and all that. And now you don't hear them saying that. So, to be honest, when I built the data cloud at ADP, Catherine, which you may remember, one of the things that I heard very carefully for me from the clients that I was dealing with was reporting and tabular reporting does what we needed to do. We're not statisticians.

[00:23:25] So, when you give us the medians and you give us the 75 percentiles and you try and benchmark this stuff, that's over our heads. So, you're overcomplicating it or you put a graph there. Why don't you put a graph there? We don't want to see the graph. We want to see the data. Because they can't. And I think HR is very slow to change in some of this stuff. Yeah. Enjoying the conversation? Don't forget to hit subscribe so you never miss an episode packed with insights and strategies.

[00:23:55] Now, let's dive back in. So, Catherine, I guess that gets us down to another path, which is, so there will be careers created out of this and there will be reskilling and upskilling. Mm-hmm. So, does HR need to lean into that now and start preparing their workforces to be able to be more nimble toward that? I would say so, but I would like to think that HR was always on top of what new things

[00:24:23] affect business and would be looking at what skills we're going to need down the road anyway. Now, do they? Meh. Maybe not. But I would want to see that. And if I were running an HR department, I would want to be able to, I mean, I used to talk about workforce planning and I didn't mean planning for tomorrow. I meant, what would you do if? Right. I think the most stellar example of that was that old one with, I believe it was Phillips

[00:24:51] in the Netherlands or one of those companies that said, what would we do if the most improbable thing happened? What was the most improbable thing? The fall of USSR. Right. Exactly. Yeah. It was Royal Shell actually. It wasn't filled with Royal Shell. And they said, okay, well, let's work that scenario out. What would we do if there was a fall of the Soviet Union? Well, they were ready, which had a big impact on their oil strategies at the time. Yeah.

[00:25:20] Because, you know, and then when you're thinking about it, well, what's the most improbable thing? Well, little men from Mars are likely to come. Well, that's not very, I don't think we should plan for that, but I think we should be able to put people in a room sometime and say, what are the worst things that would happen to our company or the most, or the best things? Right. What is the most probable and the least probable within the realm of possibility?

[00:25:47] And people never get the time to think about, you know, how everybody's day job is so busy that that seems like a, you're asking me to think beyond this week. Right. Right. Or, and when we looked at companies used to plan kind of long-term and now they plan for the quarter. Yeah. A quarter is really short in the business life. Oh, of course. But you're just planning for that kind of, you know, do we make our numbers in the quarter?

[00:26:14] But I think right now, and especially if you look at the velocity of change today, there's so much happening in the world of business and politics and the economy that trying to do a three-year plan, if I did a three-year plan, hey, listen, if I did a three-year plan January 1st, by January 20th, that three-year plan is gone. Absolute. Yeah. Okay. Yeah. That is certainly a point. Right.

[00:26:40] I mean, if we just look at like, what is the effect of tariffs on, on healthcare right this time next year? Exactly. Well, we don't make all those technical machines that are in clinicians offices. Absolutely. You know, we don't make those here. Right. And they have a lot of very complicated parts. And we don't make the parts and we don't put the thing together and we don't fix it.

[00:27:05] You know, so there's all sorts of ramifications that nobody ever thought about probably of what will happen with a tariff on just the healthcare stuff. But the reason why we did was because we made a strategic decision in the United States that we were going to cede control of manufacturing, especially of steel and other metals outside the U.S., especially to China. Right.

[00:27:32] We made that strategic decision and said we were going to be a service-based economy. Right. And we were going to import talent from outside the U.S. to help us with services and professional services. And we were going to export all this other stuff so that they could have competitive advantage in those things. We could have competitive advantage in the things we're good at. And I mean, that was the world of the economy up until, you know, basically January of 2025.

[00:27:57] Well, I mean, just look at the chip question for a minute, because most of our chips come from Taiwan. The thing, the machines that make chips are very complicated. There's one company that makes these things that deal lithography. They're only in the Netherlands. And that thing goes into something. There's chemists, there's physicists, there's all these people that are involved in making one microprocessor.

[00:28:21] Now, to build a fab plant costs $5 to $10 billion and takes at least five years. Right? So, let's say we start today, we're going to build chips in Des Moines. Right. In five years. Assuming we had the people in Des Moines who can build a fab or could ship the people there to build a fab. How many people want to build a fab?

[00:28:49] You know, they're very clean. They can't go on a fault. Right. God knows it has to be very still. So, I don't think Des Moines has earthquakes. I could be wrong. Who knows? These days. Right? They might have tornadoes. Sure. But, so, planning for that is really complex. It's not like, oh, we can make our Apple phones and somebody just needs to screw the stuff in. Right. That's not how complex technology works.

[00:29:17] Well, it's not how the vertically integrated systems have worked in companies who are manufacturing these really complex things for years. Right. You know, even the concept of a car being made in the U.S., to your point, 90% of the content is made outside the U.S. It's just, it's just put together here. Right. And so, it doesn't happen on a, it doesn't change because there's one person says make it change. Right.

[00:29:45] And the thing is that we, those decisions were made on purpose. Yes. Because it was much cheaper to have mainly women in Taiwan doing a very, because they have very nimble little fingers, doing a lot of the work on semiconductor builds. Right. And it was, they employed a lot of people. We did a lot of good for a lot of Asian countries doing stuff like that. That's right.

[00:30:11] But they, that meant that we got chips much more cheaply than we could build them here. But it's true. Yeah. But even if you move that in back to the AI conversation, where server farms were created for processing tons of data. Uh-huh. And one of the things that we found out was, what did they need a lot of?

[00:30:41] Power. Water. Water. Power and water. Power and water. Yeah. Exactly. The amount of, people don't think of, I wrote an article about that for Workforce Solutions recently that just looked at what is the cost of AI when you look at the production of it. I mean, we all know you need lots and lots of data. Well, now we've got, you know, people have been dumping the data and the big things for enough years. We have a lot of that kind of stuff, we think. Oh, yeah.

[00:31:10] But Microsoft just bought one of the Three Mile Island nuclear reactors. Oh, like. Really? They're going to put it back into service. Was that one of the ones that melted down, Catherine? It could be, but they're going to redo it. Wow. And they're going to, they bought all the power that that thing will produce. It'll be enough to cover 800 families, 800 homes could run for their lifetime on the power that Microsoft has now bought.

[00:31:41] And again, you don't just take your nuclear facility and turn it into a working thing overnight. So that will take time. I think it was, I believe it was Amazon that just bought a lot of the power from the Tennessee Valley Authority. Wow. Dude. I think it was Amazon. It might've been Google. I could be wrong on that, but I thought it was Amazon because they need that. But it also takes a lot of water.

[00:32:07] And my point in writing this was saying, dear HR people, can you do anything about this? No, this isn't. I just want you to be aware that there is a cost to what, what you're using. If you're, you know, just to understand that when we make decisions to use things that are so cute and Alexa will give you your recipe. Right. And if you make it and watch her chop peaches or whatever, that there is a cost behind that.

[00:32:38] I'm sorry. There she goes. There she goes. Would you like me to cut some peaches for you, Catherine? Alexa, stop. She is telling me what kind of peaches are the best. Sounds so pleasant when you say it. I'm never pleasant when I'm. No, I'm not either. Alexa feels like an abused partner for me. Right. But, but you're right, Catherine. And that's one of the things people don't realize there's an actual cost, especially when they're using Copilot inside of Word or Excel or whatever.

[00:33:07] But when you're developing a model or an algorithm like Dwight and I have been, you know, dabbling in that now. And when you see the actual cost of running that one statement, and there's really literally a cost associated with it. Right. Pennies add up. Yep. It's phenomenal. And somebody pays for it somewhere really quickly. Mm-hmm. We are. We are. You are mentioning it. Right.

[00:33:35] But it will, to your point before, if, if Google or Amazon or Microsoft are buying these facilities, that means those facilities aren't in business giving us that energy. Right. And so, if anybody remembers their economics lessons, if supply goes down, what happens to prices? Prices go up. They go up. Right. Yeah. Yeah.

[00:33:59] So there's an inflationary element of this that we have to take into consideration for our future, that the cost of artificial intelligence isn't just jobs. It's not just other things. It will literally be cost. It will make things go up. Like, the cost of water. Mm-hmm. And the cost of power, at least. Right.

[00:34:21] And if they're building great big sites, data sites, you know, a bunch of years ago, they were building data centers. Yeah. And they were housing cloud things. And right now, that seems trivial compared to what these builds are doing. And they're going to take farmland. They're going to take, you know, massive areas of land because they're huge now. Well, the farmland's not going to be used anymore because the farmers aren't going to get their subsidies anymore. So. Right. Right.

[00:34:52] And alternative power, like solar and wind power, they could be used for some of this. It's sort of suddenly getting out of vogue, I guess, with, you know, a coal run fab or something would be highly ineffective, I think. Well, we're going to see. Because coal's coming back. Yep. And black lung is too, I understand. That's the cost. History repeats itself. Yes.

[00:35:23] Humans repeat the obvious, Douglas Adams once said. Yeah. No matter how technologically, what's the word for it, adept we are, we still can't solve the old stuff of history. It just comes right back at us. Oh, yeah. I mean, and the things here in this world that we thought we had eradicated, like measles, you know. Right. Because people are like, yeah, I don't want to vaccinate my kid. I'm sorry. This being, this, this devolved and became a political discussion. Yeah.

[00:35:54] It's kind of hard not to, but there are some things where you just hope that human reason would prevail. Yeah. Not computer reason, but human reason. And the problem, I mean, I blame a lot of it, in fact, on the failure of the education system as an educator. Because what, when was the last time we talked people? Logic. How do you look at something and determine if something is possibly true, remotely true, totally fabricated? Yeah.

[00:36:25] And unfortunately, the misuse of AI is really confounding that. When you make a picture so clear that it looks like exactly like you doing something you shouldn't be doing. Right. Seeing something you shouldn't be saying, and it sounds and looks like you. And a lot of people who don't rely on judgment or rely on thinking things through, or is this conceivable? Mm-hmm. Um, believe it.

[00:36:55] Remember the book, A Nation of Sheep? We're definitely going to put it in the show notes so people can help it up. Yeah. But that's the, I mean, but that's the problem of, to your point, the educational systems, but it's also a problem of our society right now. Of people believing certain things over others when logic won't prevail. Right. And, um. And science is sort of off. Yeah. Yeah. I like science. I mean, I wasn't good in science particularly.

[00:37:23] I had to really struggle in high school chemistry, but I think it was early in the morning and I wasn't too awake either. But you know, I like the idea of science proving things. Yes. As a researcher, I like to see proof and understanding, given this, that this can happen or this can't happen. Or is likely to happen. Or statistically it'll never happen. Right. You know, that kind of thing. Mm-hmm. But that's me. That's us too. Yeah. I mean, we're the HR data labs for a reason.

[00:37:53] We believe in data. We believe in measurement. We believe in the science and statistics behind science. And, um, to your point, you know, you show us a picture. I won't believe it even. But if you give me the data behind the picture and you give me the data behind the hypothesis, let me draw my own conclusions.

[00:38:22] Catherine, this is so much fun. We're going to have to do this again. Absolutely. Without my gardeners. Yeah. If they want to join, maybe we can get their perspective as well. The Latino perspective on artificial intelligence. Yeah, maybe. Yeah. Yeah. But they're definitely not going to get their jobs taken. Although I have seen lots of robotic, um, mowing systems.

[00:38:46] Um, actually given ice activity here in the, in the, um, Southern California, their jobs may vanish overnight. Oh my gosh. Oh yeah. That would be very scary. Mm-hmm. I'm sorry for that. Yeah. Well again, Catherine, thank you so much for being here. Absolutely. Anytime. My pleasure. We'll definitely call you on that. All right. Thank you very much for being here. Thank you. Thanks for being with us today, Catherine. Yeah, absolutely. And thank you all for listening. Take care and stay safe.

[00:39:17] Thanks for listening to the HR Data Labs podcast. Don't forget to subscribe and share with your network. Also check out the HR Data Labs Brown Bag Lunch every Friday on YouTube. Stay safe. Hi, I'm George LaRock and I'm looking forward to exploring the critical trends shaping the future of work and technology with you over on the WorkTech podcast. Now this podcast is a little different.

[00:39:47] I bring together industry leaders, innovators, and investors, and we go deep into market intelligence that matters to HR pros and tech providers alike. So give the WorkTech podcast a listen here on the WorkDefined Podcast Network and please subscribe if you like it. See you there.