Episode Transcript
[00:00:00] Good morning. Today I'd like to make some comments about the American economy and what we're seeing in the job numbers, the productivity numbers, the data from anthropic on job displacement, et cetera. You know, I don't claim to be a trained economist, but I know a lot about the economy from my own experience in business and talking to many, many of you. And it's really interesting when you look at the charts that in the last 12 months here in the United States there's been virtually no new jobs created. Almost none. The net is virtually zero. Of course there's been lots of push on reducing immigration and reducing the size of the labor force, which presumably could create higher levels of employment. But unemployment rate has gone up a bit also because a lot of the job loss or transformation has been moving away from the more high value jobs towards frontline work, healthcare work, hospitality, transportation, entertainment. Things that have sort of illustrated, I would say, both the healthcare part of the economy and the experienced part of the economy. And you know, the consumer sentiment in the United States is at a 12 year low. People are not feeling good about the way they're economic lives are going.
[00:01:19] And gas prices are very high here in California. I was just driving by a gas station Yesterday, it was $5.70 a gallon. Now we're in a very high tax area, but I don't think I've seen it that high in a long time. Every now and then it hits $6. But that's clearly going to happen with this new war. And of course the tariff regime has been a tax on everybody, consumers and producers and exporters. So we've all been dealing with a little bit higher prices from that. And then yesterday when I was listening to Bloomberg, one of the analysts made a very good point that here in the US We've had above target inflation for about a decade. And so what's happened is we've all become accustomed to inflation and watching prices go up, assuming prices are going to go up and that devalues the money we have and that results in speculative investment in crypto in the poly market and other forms of gambling in the stock market and other speculative assets. So I personally think we're kind of going in a strange direction. And I'm not going to talk about the politics of this because honestly, I kind of wish the Trump administration had done nothing because I think the economy would have been fine without all this. But, but we're dealing with a political leadership in the United States that really likes going to war.
[00:02:44] So the parts of the economy that are growing the most are the military industrial complex, the aerospace industry, the space industry, and the defense industry.
[00:02:55] Now, you can make all sorts of arguments about whether we should have more or less guns versus versus butter in the economy. That's the way economists think of it. But the trend to me is probably not good here. And you know, you add to that the lack of global partnerships and global alliances and focus on global alliances. Now, if you think about all of that as a whole, it sounds a little bit negative. On the other hand, what I've learned over the years, over many, many years, talking to many of you and being a business person and running my own company, is that the thing that really fuels this country and many other countries is this sense of creativity and innovation and opportunity and ambition that people feel if they come here. And this is very much still true, despite what's been going on with the lack of DEI and other topics, that if they come here, they can build something, they can build a business, they can build a career, they can build a company, they can build a product, whatever it may be, and it will be relatively easy to get it into the market. And you will be glorified and celebrated and rewarded for that work. My father was an entrepreneur. He was really a scientist, but he became an entrepreneur. My mother was an entrepreneur. I became an entrepreneur later in life.
[00:04:17] And it was hard, and it is hard, but you feel a sense of empowerment in the United States that is built into the culture. The people that do gardening, the restaurant workers that open restaurants, the people that build small businesses here, immigrants that come here. It's a very empowering country. Now I live in California, where we have lots of immigration and you see new businesses being formed everywhere all the time. It's very exciting and energizing. And so even though there's policy decisions being made that might feel like they're slowing us down, the creative engine of this country is great. And that sort of leads me to the next topic, which is AI. Now, you know, I love Anthropic. I think in my own experience, Anthropic is technologically advanced beyond ChatGPT and maybe Gemini in many respects, not. Not in every area, but in certainly in the analytical areas and its code writing abilities. And we default to anthropic in Galileo. But one of the comments and the reports that just came out yesterday from Anthropic is about the job loss that's expected due to AI. And I want to talk about that because it's a really important topic for all of you in HR and All of us trying to figure out how to implement AI in our companies. Now last year we did a long project for nine months and probably talked to 60 or 70 companies about this idea that the way to implement automation is to do job task analysis, take jobs, decompose the tasks, automate the tasks, and voila, you will improve productivity. And what Anthropic has done in the most recent report, which I'll link to, is they've created a language called AI exposure. What percentage of job a, job B, job C is exposed to AI. So if you're a software engineer, 50, 60, 70% of your job is exposed to AI because theoretically it could be done by AI. If you're a typist, 90% of your job is exposed to AI. If you're a transcriber or a translator, 99% of your job is exposed to AI. Anyway, they go through that. They built a polar chart there. And when you look at the polar chart, you see that in their analysis, business people, analysts for engineers are highly exposed. Healthcare workers, entertainment workers, transportation workers, repair people are not exposed because you know, an AI can't repair your refrigerator.
[00:06:48] So you know, at first glance it makes a lot of sense.
[00:06:51] And if you applied that methodology to your company or HR, you would say about 40% of the jobs in the company are going to go away. But actually all you're really saying is 40% of the tasks are going away as they are defined today. Now the flaw with that approach, which we discovered last year when we were doing all this work, is that human beings have agency.
[00:07:20] When something gets automated, we don't sit around and twiddle our thumbs with nothing to do. We do other things. We are by nature value creating animals. And that's why human beings are different from machines. We have creativity, we have curiosity, we have ambition, we have career growth, et cetera. It's just sort of the way we're wired. And you know, I think this is DNA. I think the, the learning DNA in our bodies is maybe one of the strongest instincts that we have. It's, it's what brings us into the world as child and what brings us into the world as adults. And so when an automation thing comes along, whatever it is, and makes something that used to take a long time happen very fast, we are marveled by it for a few minutes and then we just accept it and we do other things around it. And believe me, it actually is only surprising for a little while. I think self driving cars are going to be spectacularly interesting for a little while. And then we're going to just assume they're there, we're not going to pay any attention to them. Word processors, spreadsheets, PCs, audio recording, mobile phones, automatic video creation, automatic graphic creation, all of these things that are just spectacularly interesting and fascinating for a little while become commonplace once we get comfortable with them. Because we're learning animals and we're always looking for ways to add value on top of them. That's what's going on in the market of AI. Many, many things that were quite exceptionally interesting a year ago have become commoditized and we're adding more value about them. And so what happens when you automate tasks in the jobs that anthropic would say are high exposure is the people doing those jobs do more, they create more interesting software applications, they create more complex solutions and hopefully we give them problem centric jobs that they're not sitting around looking for things to do. So we start to solve level three and level four problems in our maturity model, multi process workflow related problems. I just had this conversation last night with my son in law about his business in insurance. He does a lot of work in insurance and there's a lot of paperwork and forms and data to manage. And he's using, starting to use Claude to figure out a way to automate the work that he does with his customers. And he said, you know, by the time I get this done, then I'm going to have to go back to the company and show them what I did. And I said, yeah. And then what'll probably happen is you should figure out what the next step in that business process is so that your automation tool can be coupled to the next step in the business automation tool of his company. And his company's in the middle of implementing a big system to do this. So these high exposure jobs don't go away. If you look at software engineering jobs, they're going up. The number of software engineering jobs is going up. I just wrote an article about this. I've looked on Lightcast, very carefully, analyzed the data and there's lots of new jobs being created. And that's because the faster you can generate code, the more things you can do, the more problems you can solve, the more applications you can build and you start to operate at a higher level. You know, when I worked at IBM in the 1980s and we were writing code in Cobol, I mean Cobol is actually a pretty high level language, but the problems of, you know, writing machine language and figuring out how to access the data in the hard drive and that doesn't really add value. Those are necessary things to do. But if that can be automated and we can work at a higher level, we can solve problems in a more abstract way. And that's what's going to happen in software, that's going to happen in data analytics. That's what's going to happen in all these high exposure jobs. So these high exposure jobs are not going to go away at all. They're just going to become higher level jobs, probably higher paid, and there will be many, many more of them. I mean you can write code yourself, just tell Claude or tell Galileo what you want it to do and it will write code for you. You don't have to go to school and study computer science unless you decide that's what you want to focus on. Now the low exposure jobs, which anthropic things aren't going to be touched by AI, actually have huge exposure to AI. Look at a nurse. Okay, so I just went to the doctor the other day. The waiting room was filled with people and they cranked us through in a big hurry because they're so well automated in their scheduling, in their record keeping, in their diagnostics, in their prescriptions. I've had a cough for a few weeks. So I go into the doctor to take a look at my lungs to see if I might have pneumonia because I've had pneumonia before. And the doctor listens to my lungs and they weigh me and all that stuff. And she's great, she knows what she's doing. She's typing into the little thing there. It's all very fast. And I said, you know, just in case, maybe I should get a chest X ray. She goes, okay, let me call them. She calls the chest X ray place across the hall. She says they're available until 4:30. Let's send you over there. Boop. Presses a button, a little piece of paper pops out. I walk over there, the guy goes, oh, Mr. Burson, welcome. Come on in, let's go in here and do this. I was out of there in 20 minutes with the whole thing. Now I'm not no if. I don't know if that's AI, but AI is making that job way easier than it used to be. They used to fill out a pieces of paper with courier. There used to be a little vacuum courier thing in the hospital. Do you remember this little thing? They'd stick your paperwork in a tube and it would, some vacuum suction thing would suck it out of the room you're in and suck it to the room you're going to. I mean, that's all gone.
[00:12:49] So AI scheduling is massive in frontline work. Baristas, coffee people that work in coffee shops, people that drive trucks, the Amazon delivery people, they're all getting productivity improvements from AI. Let's talk about repair people. I mean, it's ridiculous that Anthropic thinks AI is not affecting repair. That's completely wrong. I know this. I studied the repair space years ago when I was doing training industry stuff. And I remember, for example, in British Telecom, and this is a really interesting story, the British, you know, country has many, many, many old phone systems, as does the United States. So you've got a phone that's broken in your house or in your business, and somebody from BT comes out to fix it, and they go in the closet and they look at what you've got and they say, I have never seen this before. I don't know what the heck this thing is. So I got to go back to the office and figure out what you've got. And, you know, a week later they try to solve it. This happens to me, by the way, with our AT&T phones all the time. And the phone company hates this. They don't want to spend all this money on this. You're not paying that much for it anyway. They'd rather you use digital or some new wire thing. So they actually would like you to just dump it, but you don't want to dump it because you like it and it's all wired into your house. Well, what they did at BT is they created at the time a system where you could take a photo of the equipment and send the photo to a social network and somebody in the company would look at the photo and figure out what it was. Well, now you can take a photo and send it to chatgpt. And chatgpt knows what it is. I did this with my own router. It was so funny. I was installing a new router in our house and a little confused about the settings. And I sent a picture of the router with the model to chatgpt in a deep, quickly, almost immediately told me exactly what to do to configure it. I mean, that's a massive improvement in repair and service and extend that to all the different repair and service types of things people do. So Anthropic missed that. They completely missed that idea. Now, I'm sure the guys at Anthropic are brilliant economists, but they don't necessarily understand the way jobs really work because they're looking at the Tasks. Now, job task analysis goes way back, long before I got involved in it.
[00:15:05] And I would imagine the reason we did or do job task analysis goes back to the industrial age where most of the jobs were manufacturing labor types of jobs or agricultural types of jobs. And there were a lot of tasks and there were some tasks that some people were good at and some tasks that other people were good at. And then if we broke the job into tasks, we could better decide who could do which task. Like if one of the tasks was picking up a 50 pound box and moving it from place to place, we're not going to give that to, to a hundred pound young lady who isn't strong enough to do that. We're going to give that to the big, strong, tough guy. And so the task analysis gave us the data to decide what skills were needed and who could best do what part of the work. And that goes back to agriculture and transportation and railroads and all those kinds of things. Today, when many, many of the tasks are not physical, not all of them, but a lot of them are not. I mean, even working in a Starbucks is somewhat physical, but there's a lot of automation, that kind of analysis isn't really that useful. What's more useful is the business process analysis of what is the value chain of this work and where is value added and therefore what capabilities and skills are needed to add value. And if you assume that human capital is a learning entity and you assume the human capital beings will adapt to the environment, then automation gives the opportunity to improve scale, productivity, quality, time to market speed, customer service, all these very high value things. Eliminating tasks is almost irrelevant. And you know, I think a lot of the AI push that people are getting from their CFOs or even the block story that came out the other day is using this idea that oh, if we implement AI, 40% of the jobs will go away and we can get rid of 40% of the people. It's not going to work that way. And we see it already. We did a lot of work on HR and we have this wonderful blueprint which we're not publishing by the way. It's available for clients or for advisory. And you can go into the HR department and you can look at the 94 capabilities of HR and you can literally run an analysis in Galileo and you can see what percentage of each of those 250 job titles could be automated. And you could, you know, total that up in a spreadsheet and say, therefore, let's get rid of 40% of our workforce in HR.
[00:17:36] Okay, so the HR department's 40% smaller.
[00:17:39] And yeah, maybe that saves some money, Certainly saves some money. But does the company benefit from that? Maybe not, because that isn't really the problem. The problem in employee service, for example, is employees want to get answers to a question quickly and easily wherever they are. And they don't want to have to call around and find out which HR person to talk to.
[00:18:00] And we don't want an HR team scrambling around looking for data that's out of date and trying to run reports when they don't know how. So if we automate the employee service, HR business partner role as much as we can, the employees are going to get faster service, they're going to get higher quality service. And the people that were scrambling around doing tactical work are now available to do higher level work. There's plenty of things like that to do. What about working with a manager on how to improve the productivity of his team? What about training a leader? What about building a succession plan? What about analyzing skills? What about doing workforce planning? What about looking into the job market? What about competitive analysis? I mean, we can do all that in HR if we don't have to do that tactical stuff. So even if the job task analysis does say that you could eliminate 40% of the jobs, you're probably not going to eliminate them.
[00:18:50] Now, you know, doing that job task analysis is useful, but for re engineering it's also yet less useful than you might think. Because what's going to spit out of it, and I've seen this many, many times, is you're going to see a list of tasks that you already knew needed to be automated. It's not going to surprise you at all. You're going to see tasks like analyzing data, looking up compliance rules, writing answers to questions, et cetera. You're going to see tactical tasks. When you look at the output of the job task analysis and you're going to say, yeah, you know, we've been talking about automate that, automating that for years, but we just never got around to it. So what the job task analysis will do is it'll give you tricks and tips and ideas on how you want to reorganize work to streamline some of the bureaucratic things that are getting in the way. I mean, in the sales organization and HR is really very similar, but sales is an easier one perhaps to understand.
[00:19:51] You or a recruiting is the same way. You have this process that starts with a lead or a customer or an inquiry or a job wreck or a candidate, and then it goes through this chain of activity from fitting the candidate to the job or fitting the lead to the product, assessing the strengths and weaknesses of the candidate, assessing the strengths and weaknesses of the, of the opportunity for the salesperson, refining the fit, deciding how to respond to the fit, in the case of recruiting, interviewing and assessing the person and so forth.
[00:20:27] And if you didn't have automation, you would be manually moving data through that chain, through documents or conversations or meetings. Well, the more integrated that becomes, the more that data flows back and forth across that process.
[00:20:42] And in the case of recruiting, we built ATSS kind of to do this, but ATSS weren't workflow systems, they were databases.
[00:20:49] So the ATS does some of this. In the case of sales, the CRM does some of this, but not all of it. Until more recently, now that we have much more AI in these CRMs. So you can sort of visualize in both cases or any, any case, including L and D and employee services and so forth, onboarding all of them, how these steps could be more integrated. And the job task analysis will help you do that. But honestly, you already know a lot of this. So I don't think the work that Anthropic did is as profound as you think. Because what we found last year is that when the companies who did job task analysis looked at the data, what they discovered was that the job titles and the job definitions were institutionalizing the unproductive processes they had. Because what we've done over many years, and this is a normal sort of behavior in companies, is as a process expands, we break the work into more and more jobs. So if you have 10 people in the salesforce and all of a sudden you go from ten to a hundred, you're not going to have a hundred people doing the same thing. You're going to take of the hundred, 10% are going to do lead gen, 10% are going to do qualification, 10% are going to do first call scheduling, 10% are going to do proposal generation, et cetera. You're going to break them into task oriented jobs. And then when a technology like AI comes along, you're going to look at all those jobs and say those jobs are all going to collapse. And so the process of actually improving performance isn't eliminating just the tasks and the jobs, it's redesigning how the process works. And that's what's missing in the anthropic analysis. And the reason I mention anthropic is because I'm so enamored with them at the moment. And when I see Dario saying things like 50% of jobs are going to go away. He's, he's actually wrong. That's not what's going to happen. 50%, in fact maybe 80% of jobs are going to change, but there will be many, many new jobs that are higher value that will be created. Now if you're a graphic designer or a writer and you really don't want to do anything else and you love creating little graphic objects and art and stuff, or you really like typing, or you really like grammar, you're at risk. I hate to tell you this, but that is automatable. The women in the steno pool at IBM when I was a young guy are not doing steno pool stuff. In fact, it's funny, I was thinking back about them. The way it used to work is you used to take a piece of paper and write a document by hand, in script. You would put it into the inbox of the steno pool and they would type it up for you and then you would go get it as a typed up document and then you would mark it up and take it back to them. And the women that worked in there, it was mostly women, were very nice, kind, service oriented people and they, you know, are having a good time typing stuff up and talking about things and you know, they knew a lot about grammar and punctuation and things like that. All those jobs are gone. I would venture that none of those people are unemployed. They all went into customer service, they all went into executive assistants, they all became salespeople or who knows what, that's what's going on. So that's the job recomposition stuff that's going on in the economy.
[00:24:05] So getting back to the beginning where I started this, if you look at the US data, where it's the easiest for me to analyze, this last month, for the first time in many years, I think at least 2, 3, 4 years, the number of healthcare jobs went down. And I think maybe what we're beginning to see is, is the effect of streamlining and automation in healthcare. Now the healthcare industry in the United States is very complex and very problematic as it is everywhere else. And those are tough jobs that have rapidly changing technologies and science and diagnostic tools and all sorts of things around them.
[00:24:43] But the reality of that exposure analysis for anthropic, at least the way I see it, is that we're all exposed to AI. And this is the wonderful nature of this industry, is that we are going to see reinvention like we've never seen before. Meta is now making A big deal about AI in our glasses. Now they're using it for sort of maybe spying and surveillance purposes. But, you know, you're going to be able to put on a pair of glasses, look at a situation, or if you're a repair person, a device, and the glasses are going to tell you what to do. And that's not going to be 10 years from now. That's going to be one year from now. You're going to be a nurse or a repair person or a salesperson. You're going to have a question and you're going to either say it to your phone or you're going to type it into a chatbot and you're going to get an answer and you're going to have a Galileo behind the scenes helping you, and then you're going to have a learning platform helping you, and you're just going to get an answer. So this idea of dynamic enabling us to do our jobs better is massively transformational for all of us. And what we need to do in HR is we need to move swiftly down this value curve, find the technologies and the vendors that understand the problems that we're trying to address, and they're going to take some time to figure out where they want to focus. Because right now, a lot of it's still raw technology. And we're going to transform our economy and our jobs and our companies relatively quickly. And under the covers of that is this idea of ambition, growth, opportunity, human centered leadership, learning, development, and a lot of curiosity. So even though the US Economy might be stuttering at the moment, and we have a war going on and all sorts of other, you know, difficult problems to discuss in the country, this is going to go in a positive direction. And all I'm trying to do in this podcast today is just give you a sense of how we see the Super Worker organization and the Super Worker effect on the economy and what we think you can do to take advantage of it. I think I talked in the last podcast about what's going on here. I'll see you guys at Unleash, at Transform. I'll see you guys at HR Tech Europe. Come to Irresistible in June and we got a ton of stuff to share with you on what we're working on. That's it for now.
