As the development and implementation of generative AI (GenAI) continues at a rapid pace, the universal question being asked is, “Will my job be replaced by GenAI?” In this episode of All Things Work, Gad Levanon, chief economist at the Burning Glass Institute, joins host Tony Lee to discuss topics including whether GenAI will replace jobs, its potential to transform work across industries, and how HR leaders can prepare for workforce changes.
As the development and implementation of generative AI (GenAI) continues at a rapid pace, the universal question being asked is, “Will my job be replaced by GenAI?” In this episode of All Things Work, Gad Levanon, chief economist at the Burning Glass Institute, joins host Tony Lee to discuss topics including whether GenAI will replace jobs, its potential to transform work across industries, and how HR leaders can prepare for workforce changes.
Music courtesy of bensound.
Rate/review All Things Work on Apple Podcasts and Spotify.
Tony Lee:
Welcome to All Things Work, a podcast from the Society for Human Resource Management. I'm your host, Tony Lee, head of content here at SHRM. Thank you for joining us. All Things Work is an audio adventure where we talk with thought leaders and tastemakers to bring you an insider's perspective on all things work. Today we're discussing generative artificial intelligence, GenAI, and its impact on the workplace. Experts say that unlike other forms of automation, GenAI won't just impact blue collar and entry-level jobs, it will redefine many highly skilled professional jobs and disciplines across most organizations.
A newly released report from SHRM and the Burning Glass Institute titled Generative Artificial Intelligence in the Workforce found that the accelerating adoption of GenAI will cause the broad transformation of jobs such as financial analysts, lawyers, software developers, marketers, and of course HR professionals. While few of those jobs will be fully replaced by GenAI, many will be augmented and fundamentally reshaped by the rapidly evolving technology. Joining us today to talk about GenAI as Gad Levanon, Chief Economist in New York of the Burning Glass Institute, which he joined two years ago after spending 18 years at the conference board where he was founder of the Labor Market Institute. Gad, welcome to All Things Work.
Gad Levanon:
Thank you very much. Great to be here.
Tony Lee:
Well, we're very pleased to have you join us. So, let's start with the question that many in HR hear from employees, "Is GenAI going to replace my job?" What should HR be telling employees when they ask that question?
Gad Levanon:
Well, there is the real answer, and I don't know if that's the answer they want to give.
Tony Lee:
Go for it.
Gad Levanon:
I think the real answer is that, yeah, many jobs will lead to workforce reductions, whether they'll be completely automated or whether they will increase the productivity of existing work so much that they will just won't need as many and they will cut off some of the workforce. But having said that, a lot of the reduction in workforce is not going to be just by layoffs. You can simply stop hiring, for example, and let attrition, natural attrition do its work, so not all of it will be layoffs.
Tony Lee:
Right. I thought it was interesting in looking at the report, one of the comments is that there will be this surge in productivity and output that will be created by GenAI that over time is unlikely to meet the corresponding growth and demand for goods and services. So, basically what that means is a lot of industries will be overstaffed, and so senior leaders are going to have to come up with mitigation strategies like hiring freezes to minimize its disruption. Is that where you see it going?
Gad Levanon:
Yeah, I think that will be the case in many types of jobs. And it's not starting just now. If you look at the past couple of decades, we also had significant technical development and automation and what you see that the types of jobs where there is a lot of replacement of workers by technology, you see no more hirings or much less hirings than before, and the occupation is just getting older. So, the average age, because there's no new workers entering the occupation, so it's just getting older. And I expect to see that a lot of the white collar jobs that now have a lot of young people, they will also start getting older.
Tony Lee:
Mm-hmm. I mean, the other analogy that I have heard is if you think about robotics in factories, I mean prior to robotics, you had factory workers who were handling every aspect. Now you have factory workers who handle robots. They manage the robot, so the jobs don't necessarily go away, they just change, right?
Gad Levanon:
I would say you'd have some of the previous workers in the factory who now manage robots, but the idea is that robots are cheaper than humans. So, you do let go of some of the humans, unfortunately, and that is likely to happen, but as in previous automations, I don't think the end result would be mass unemployment, because at the same time that some jobs are going away, there is a lot of new jobs that come on, and eventually the number of employees depend on the overall demand of the economy, and that's likely to remain robust, yeah.
Tony Lee:
Right. No, true. So, let's focus a little bit on HR. The saying that I love is that AI likely won't replace HR professionals, but HR professionals who understand AI will replace those who don't. Do you buy that?
Gad Levanon:
Yeah, I think it's very much the case. I would say that there will be, especially the most, I would say repetitive or clerical jobs in HR are likely to be exposed to Generative AI as well. But I definitely think it's important to understand the impact of Generative AI, and it has important implications for HR leaders. For example, one topic is re-skilling. You'll have a lot of need for re-skilling both the jobs, the workers who are being exposed to Generative AI. Are you just going to let them leave the company, or some of them you can re-skill and let them try other positions in the company. And there will be also a significant increase in demand for people who can work with AI, who can program and use Generative AI technology. And those skills, for the most part, they are now, and you would need to re-skill those people as well.
Tony Lee:
Yeah. Prompt engineer is one that seems to be jumping up training people how to tell AI what you want, right?
Gad Levanon:
Yeah. That's an example. Yeah.
Tony Lee:
And also in HR, talk about that a little bit more. We've started seeing examples of AI being used. The one everyone hears about is creating job descriptions. That seems to be one that's becoming fairly widespread, but there seem to be others. A lot of the more manual aspects of HR, I mean, I've heard of a company that used to have someone dedicated to responding to job candidates. When a job candidate would apply, they would respond back with information, and now that's completely outsourced, so to speak, to AI. Are you hearing of other examples like that?
Gad Levanon:
Yeah, I think one of the greatest advantages of Generative AI is that it's able to communicate and, especially relatively simple communications, and they can do it on a large scale at a low cost. So, what you mentioned interacting with job applicants, perhaps automating benefits services, like there are people who have questions about benefits. If before, they used to talk directly to someone from HR, for the most part, that could be automated and let the algorithm answer your questions. So, there is a lot of this low level communications that is probably going to go to AI. Conversations like, "I want to get promoted." Those, or someone was breaking rules or laws in the company or things like that, that would remain with HR. But a lot of the simpler communication is likely to go away. And another example is a lot of HR is doing a lot of surveys to get the pulse of the companies workers.
Now in most companies, there is a dedicated survey team that only they know how to use the survey software and only they can do surveys. But I think in the future, one of the big things is that a lot of the software and other tools that you now need a lot of training to be able to use, you'll be able to use in the future. You'll be able to use it with plain English without much specific knowledge or just tell it what you want to do and it will do it for you. So, I think a much larger number of people will be able to do surveys, for example, and you won't need a dedicated team to just do that.
Tony Lee:
Yeah, no, that's a great example. And there's been so much interest in employee engagement and measuring engagement. The other area that I'd say over the last five years has just exploded has been HR analytics. And a lot of companies have hired full-time analytics people in the HR department to measure everything from cost per hire to turnover rates, retention, and what have you. Can you see AI replacing some of those people as well?
Gad Levanon:
It could replace in some ways the more simple tasks, but for the most part, I think asking the questions would remain the human's job, because one thing the Generative AI is not very good at is in creativity. That's something they're still lacking for the most part. So, to know what to ask probably in the near future will remain a human task. But I would also say that Generative AI is likely to be able to create much more data on the company, at the company level, that will provide much more potential insights for HR analytics people to work on. Every single related to language or correspondence or any interactions can now be quantified somehow and improve. There are probably ways to better gauge employee engagement or retention risk using all kinds of language that is being used around the company. So, I can see that providing more jobs for the HR analytics professionals.
Tony Lee:
Yeah, that makes perfect sense. And let's loop back around a little bit to the re-skilling idea that HR is going to be responsible for helping employers. They don't necessarily want to let people go. They want to re-skill the people to continue to be productive. Where is that in the process? Is it too early yet to see re-skilling happening because of AI, or are you actually seeing examples of it already?
Gad Levanon:
We haven't seen, I would say yet major layoffs just because of Generative AI. I think it is something is probably near going to happen in the future than now, ChatGPT just started a year ago, and it will take time for all those softwares and tools to be developed. But I think it's a good muscle to have to be prepared for when it does happen that you have pathways to try to think ahead what jobs are likely to be exposed and what types of jobs you'll have workforce reductions. What are the skills that those people have? What are other positions that have a kind of similar skill base to the ones that are likely to be exposed? And how do we train those people to go up or laterally to other positions in the company? I think it's very important to develop this capability so you're ready when you do need to do it.
Tony Lee:
Yeah, no, of course. So, I can't not mention the skeptics out there who continue to point to, and there are issues. There's no question GenAI has some issues, hallucinations and all the way to copyright protection issues. There are quite a few, but I don't hear them quite as loudly as I had been. I mean, is it safe to say that companies need to start finding ways to capitalize on GenAI pretty soon, or risk falling behind their competitors?
Gad Levanon:
I think so. I think at this point, there are a lot of technologies that you, at the beginning, you don't know if it's hype or real. And in many cases it ends up as hype. I think Generative AI is not one of those. I think it's a very big deal and it'll impact every company. Some companies will be... Another thing we have in this report is we look at the type of jobs that are more likely to be exposed, and the companies that are most likely to be exposed are in the financial insurance area, because they have a lot of clerical jobs, a lot of even white collar jobs that are relatively simple to automate with Generative AI. And I think they will experience the biggest impact.
Tony Lee:
Yeah. I mean, the report mentions financial analysts, actuaries and accountants, auditors and compliance officers. So, all within the financial realm. And then on the technical side, software developers who are creating code that GenAI can create instead. And then of course, administrative roles everywhere, right?
Gad Levanon:
Yeah.
Tony Lee:
So, back to an HR perspective, a lot of HR professionals rely on their vendors for guidance and help, and it seems that GenAI tools are being integrated almost as a standard feature now into vendor technology platforms that HR is using. Would your guidance be that HR pros can rely on their vendors to help them understand and offer up GenAI in what they're doing in HR? Or should they be more proactive than that?
Gad Levanon:
Well, we are seeing the trend of outsourcing to vendors is not new, and it's been growing for many years. We are seeing since the beginning of the pandemic a big jump in that. I think partly the shift to remote work and the kind of deepening of the digital transformation and now the Generative AI, they all require relatively advanced and specific talent that most companies don't have. So, I expected this trend to actually accelerate that companies will rely more and more outside vendors than on their own talent, because it's just becoming more and more complex and companies don't have the comparative advantage in those skills.
Tony Lee:
Yeah, and the vendors are happy to get the business. I mean, they're adding GenAI expertise very quickly, right?
Gad Levanon:
Yeah, but I think the results are a big competition among vendors. Some of them are faster and some of them are slower in adopting Generative AI, and that's becoming a more complex market as well. So, everyone is feeling the heat.
Tony Lee:
Yeah. So, let's talk about the control of GenAI within the company. I mean, do you think there should be an individual at each organization who kind of oversees GenAI usage across the company, or do you think it should be more distributed that HR manages its own use of GenAI and accounting manages its own use? Have you seen any case studies there?
Gad Levanon:
My guess is it will become such a widespread technology or tools that every department, every part of the organization will use it as kind of a second nature in several years. But there probably will be centralized functions within the organization that build those tools and interact with vendors and are responsible for the overall AI policy of the company. So, it will be partly, I think, distributed in partly the more sophisticated, strategic parts of it will remain central.
Tony Lee:
And policy is an important word here. I mean, we are seeing companies creating GenAI policies for how it's used, which actually goes to the question also of should GenAI use be centrally managed, or should every employee be given the opportunity to experiment with GenAI in their job to see if they can come up with ways to become more productive?
Gad Levanon:
Personally, I found that almost life changing, I'm exaggerating a little, I think definitely became more productive in some elements of my job. I think other workers should enjoy the same benefit. It's partly personal. Some people like writing emails and some people hate it, and they are kind of more open to letting AI do the job for them. So, I think individual experimentation could be very useful here.
Tony Lee:
So Gad, as an economist, I don't see your job title on the list of jobs that may be impacted by GenAI, but perhaps it is. How has GenAI impacted your work as an economist?
Gad Levanon:
Well, so I'm both an economist and an immigrant. I don't know, you can probably tell by my accent.
Tony Lee:
Yes.
Gad Levanon:
So, I think one of the, as a non-native speaker, Generative AI has been a blessing for me, because in all kinds of communications, I write a lot of LinkedIn posts and other things, they make it slightly better and more correct grammatically and so I use it in writing all the time, but it's more than just writing. It's also in getting information. But when you ask in most questions that related to economics, you can ask and you get pretty good answers. And even ideation sometimes I kind of tell ChatGPT, "This trend is going on right now. What do you think are the implications of it?" And it would usually give you five, six implications that let's say three of them are not good, three of them are good, and two of the three you haven't thought about. So, that's already a contribution. So, I found it extremely useful.
Tony Lee:
Yeah, that's excellent. So, that's going to do it for today's episode of All Things Work. A big thank you to Gad Levanon for sharing his insights on the future of GenAI. Before we get out of here, I want to encourage everyone to follow All Things Work wherever you listen to your podcast, and also listener reviews have a real impact on a podcast visibility. So, if you enjoyed today's episode, please take a moment to leave a review and help others find the show. Finally, you can find all of our episodes on our website at SHRM.org/podcasts. Thanks for listening and we'll catch you next time on All Things Work.