Chapters
00:00 Introduction to AI in the Workplace
01:56 Understanding Productivity in the Workplace
05:09 The Role of AI in Enhancing Productivity
09:52 AI Adoption Across Industries
15:10 Empowering Employees with AI
20:01 AI's Impact on Healthcare Productivity
25:13 Measuring Productivity in Services
27:33 The Imperative of AI Adoption in Business
30:50 Historical Context of Technology in the Workplace
35:11 Navigating the Challenges of AI Integration
39:30 Empowering Employees in the Age of AI
43:47 The Role of Leadership in Technological Change
51:08 Opportunities and Risks of AI in Business
Welcome everyone to the Humans and AI in the Workplace podcast. Over the last few years, it's become clear that artificial intelligence, AI, is one of the most impactful and disruptive transformations in the workplace. As a leader, you may be wondering how to get started and how to do it in an intelligent way. Or you may be stuck on how to overcome some of the people issues and human bottlenecks your AI has crashed into.
We are here with Dr. Debra Panipucci and Leisa Hart and a special guest, Stephen King, Productivity Commissioner to discuss the productivity promises of AI. With the rise of AI in the workplace, it's inevitable that there's been an almost meteoric rise in expectations of significant and immediate productivity gains. Today, we're excited to be speaking with Stephen King Commissioner at Australia's Productivity Commission and a Professor of Economics at Monash University. Since joining the Productivity Commission in 2016, he has worked on a range of projects, including the series of research papers released in February 2024 this year on the implications and opportunities for AI in Australia. For those not entirely across the Productivity Commission, it's the Australian Government's independent research and advisory body.
on a range of economic, social and environmental issues affecting the welfare of Australians. Its role expressed most simply is to help governments make better policies in the long-term interest of the Australian community. So welcome to the Humans and AI in the Workplace podcast, Stephen. Thanks for having me. We're excited to have the opportunity to unpack this big topic with you today so that our listeners have an insight into the work of the Productivity Commission and their position and recommendations on AI in the workplace and what we can reasonably expect in terms of productivity over the short term and the long term. And we're hoping to sort of unpack some insights for leaders in the workplace as well today as part of that to get a better understanding. Sounds good. Yeah, great. So we're going to kick it off with a pretty basic question, which is what is productivity in the workplace and how is it currently kind of measured and understood in the work of the Productivity Commission?
Yeah, it's one of those questions that if I have trouble answering, I'm immediately fired. So productivity, that's right. I'd better be productive. Productivity is just simply a measure of how much output you get for how many inputs you put in. So if you think about labour productivity, which is the most common one, it simply looks at how much output you're getting from a business, from a sector or from the country compared to the number of hours people are working to get that output. Think of a banana farm, you'd be looking at how many tons of bananas you get out and how many worker hours you need to get those tons of bananas. There's other measures of productivity that go broader than that which take into account other factors such as most obviously capital, different forms of machinery. Labour productivity is certainly the one we care about because labour productivity drives wages and over the longer term drives our standard of living. And we've got a high degree of services in Australia, don't we, as service-based industries? Is that a better way of describing it? Yes. So Australia, as with all developed countries, just seen this radical change due to technology, driven by technology over 150, 250 years. So if you take Australia, for example, around 1900, about 20 % of our workforce was in agriculture. If you look today, you're looking at about 3%. We sometimes have a rather glorified view of the old days when everyone had decent manufacturing jobs. Now manufacturing is a tiny part of our economy, of our output.
But we've got more manufactured stuff today than ever before in our history. So if you're looking in our economy today, we are a services economy. 90 % of people, 80 % of output is services output. And if you think about it for a few seconds, it sort of starts to become obvious. Just work through your day, what you really want, you get. You go down to the shop, buy a cappuccino. Well, yeah, you get a little bit of coffee, but most of it is the service you need to go and see a doctor, you need to go and visit your parents in the aged care facility. I mean, these are all services and particularly important of the human services like education, health, aged care, which are the biggest growth in the services sector and also have the highest level of government involvement. Yeah. Yeah, great. Thank you. Next question.
In February this year, you and the team of the Productivity Commission released three important research papers under the overarching title of making the most of the AI opportunity. In research paper one, is titled AI Uptake, Productivity and the Role of Government, it makes reference for AI having the potential to address some of Australia's most prominent and enduring productivity challenges and skills and labour gaps and productivity and services, which we're just talking about.
Apart from the obvious benefits of being able to compete globally as a nation through the uptake of AI, why is it important for Australia to embrace AI in the workplace? So it comes back to that services economy. GPT doesn't just mean the technical term, which I won't try and remember, T stands for transformer. That's the only thing I can remember of part. GPT in economics means general purpose technology. General purpose technologies like the lifeblood of technological growth of improvement in living standards, in improvement in business productivity. So think of electricity. It's a general purpose technology. You don't sort of just think, yeah, I've got electricity now so I can do one extra task a bit better. You say, I've got electricity now. The world has changed. And if you look at the pre-electricity world, you know, remember the world of
gas lights, oil lamps, then once electrification had moved through the economy, so steam engines, for example, had disappeared from factories and you had much smaller factories and much safer factories because electric motors aren't as big as big steam engines. Now that process, by the way, all took about 50 years. I mean, this was an overnight, it wasn't sort of Hey, we've discovered electricity. Wow, this can make big differences. It takes a long time. AI is like that. It's part of the digital revolution. So we need to put AI in context for all the other amazing things that we've seen over the last 50 years and are still occurring. So computerization, personal computers, is another general purpose technology. Took about 20 years. It was faster than electrification, but it took about 20 years to get for real benefits from that is also a general purpose technology. It's not new. The term itself goes back to 1956. So we've had slow improvements in AI. It's accelerated. You get this exponential growth and we're seeing that in AI. Remember, it's less than two years since ChachiPT3. Yes. It's less than 10 years since AlphaGo.
I mean, you sort of think, well, that was ancient history when you had a machine be a player, a world champion in Go, but it's really recent. And what it's going to be able to do, its big impact is going to be on services and services are the bit of our economy. And again, particularly those human services that I mentioned before, health, education and so on. They're the ones where we've seen the slowest productivity growth traditionally.
And here we have a technology that says we can suddenly make big gains in those service technologies. most of your listeners, well, all of your listeners will be users of service. If they're listeners, even if they're thinking of running their own business, hey, I'm in a business that produces bananas, I mentioned before. They're already using AI. It's already transforming their business. They may not even know it. We're going to see that sort of transformation accelerate, it's going to take time, but it's going to mean, again, just better living standards for all of us. Yeah. You know, it's really interesting when you say that, because what I've heard and seen is I get the sense that agricultural industries are adopting AI and technology and using it far greater and far better than corporate offices.
Corporate officers feel like it's just hidden in the vendor software that they've got and they'll turn it on here and there, but they're not really strategically thinking around, like for instance, a cattle farm will track their cattle and use it to monitor for a whole range of things. But I get the sense that it's not happening so much in the corporate realm. Yeah, I think in the corporate realm, you've, as you said, you've got that background adoption of AI. So any businesses who are using cloud-based software, using any of the main business programs, they're using AI in their business. Sometimes called ubiquitous AI or shadow AI. Yeah. Yeah. It's in the background. they sort of say, not my business. Well, if you've got predictive text using AI, if you have any mapping software, basically if you're doing logistics, you're using AI now. Businesses have
tended to the gains for a lot of business haven't been as obvious as they have been in say the mining and the agricultural areas where just the mix of AI image recognition and the ability to process and drones all of a sudden this has just meant huge gains already in those industries from being able to use those technologies for a lot of everyday businesses, businesses producing services. They haven't seen
the obvious uses, but they're there. What I'd suggest to your listeners is step back and say, what am I currently doing which AI can help me with? And I'd probably start off by saying, what is the boring stuff that I'd really like to get rid of in my day-to-day activities? And can AI help me there? because we're at the beginning of a revolution and AI is really, really good at doing boring stuff. And it's thinking about the boring stuff and what is it that is just using unnecessary headspace, brain capacity at the start of the day or at the end of the day or during the day that the technology can step in and help work for you. We often talk about it in the context of when we're working with leaders talk about.
mindset and have you have a mindset and encourage your people to have a mindset of the technology is here to serve you. It's not the other way around. So if you can think about it in that context, you're thinking about, what can I get it to do for me today? What could I get it to improve upon that it did yesterday? Because it will, it's at its infancy or we like to say it's the least sophisticated it is today, then it will be tomorrow, then the next day, because it will continue to learn and evolve. I think the challenge is leaders understanding that they have to create the space for their people to be able to sit back and think about that and then be guided through that. So learn how to work with the technology in a way that is starting to go actually, okay, I can see some benefits and have that mindset that it is here to serve me and it is at its most unsophisticated right now and it will grow over time talked about it before with a previous guest about thinking about it as it's like an intern. You you've to help it understand and then it'll fly over time. So we often work with leaders around managing their expectations about the journey that they have to help their people go on to be able to adopt the technology and to start to understand what it is and how it can serve them. Yep. yeah.
Interns are a good word. was sort of thinking myself of infant, but that raises, it'll grow up to be the grumpy teenager. That's still a possibility. it's, it's partly giving license to your employees to be able to experiment and saying to your employees that I'm open as someone higher off the chain, but boss for want of a better word to you're doing things differently, you will be evaluated on the outcomes. I'm not going to say this is the way we've done it for 50 years and damn well we're doing it for the next 50. Businesses that have that mindset will die. And I'll give a PC example, know, so because it is a great example. There are so many places, particularly the younger staff members at the Productivity Commission already experimenting around, often in the shadows, and I'll explain why in a minute, but we get literally tens of thousands of pages of submissions which have been manually read and documented so that, this is actually relating to this particular issue, so I must now notify the person about that issue, and it is one of the most boring jobs that you can possibly imagine.
That can all be done by AI so that that first initial read through and categorization of where's the material relevant for what so that the people who are doing the research much easier for them to come back and sort of say, yeah, this submission is really important and this bit is really important for me. And that's just a huge time saving. AI research assistance. This was an eye opening moment for me had to, for various reasons, look into New Zealand pharmacy rules. And I was thinking, yeah, five hours of my life has just disappeared. I started using an AI research. I said, let's see if this can make it much faster. And I had the answers in 20 minutes. And I'd been able to cross-check them because... Perlucination? Well, and...
The good AI research assistants actually give you the references they've got the information from and it's literally a click and yep, that's what that web page says and it's a government web page that's reliable and so on. So the sort of checks that we all do when we're doing research and doing internet searches, for example. Yeah, I save four and a half hours and that was great. I probably wasted it watching YouTube or something like that.
But there's these huge gains. Just give your staff the freedom to experiment. Give them that license. Make sure they understand, I'm not going to evaluate you or judge you on how many hours I see you sitting down at the computer typing away or anything, that old school stuff. I care about the outcomes. You are responsible for the outcomes. And I think that...
You know, that responsibility bit is something we forget about often with AI, but you have the license to do that. Now, why is the Productivity Commission a great example of this? Why do they say often this is occurring in the shadows? Because we're a government agency and the government has been more rapid than I actually expected in freeing up the ability for agencies like the Productivity Commission to use AI and it
I think it was literally yesterday came out with a new set of rules and guidelines, but it's still a conservative organisation. so the government is an example of an organisation that often focuses way too much on how things are done rather than getting what the results are. And I think government also needs to change that mindset. And there's a part of that which is about risk, managing the risk of introducing new technologies without having thought through what the guardrails need, what the policies need to be put in place. And we always talk to leaders about that too. So to your point, encourage your people to experiment, give them the space to experiment. And we would add to that, give them the guardrails to play in so that they feel safe to play in terms of what the policies are and the guidelines without bolting everything down where you just kill.
any kind of creativity or innovation or playfulness or experimentation. So give them some guardrails to play within, but let them go and encourage them. But to your point about the government, they're more risk-inverse, which possibly slows down their ability to give people an opportunity to play more broadly. Is that fair? Or you think it's just... Your guardrails point's actually really, really critical.
I don't know, one of our major retailers who decided they were going to use a lot more facial recognition and just saying, yeah, somebody was given a little bit too much latitude there because of that decision. Clearly had never got to the point of all the level we used to be saying, hmm, okay, what are our customers going to think about this? Is this, you know, is this going to satisfy the front page test. The pub test or your mum test or anything. Yeah, as I'm saying front page test, mean that ages me. Front screen. In previous roles where you're making important decisions that have public implications, the question always was, are you happy to see this on the front page of a Daily Telegraph or for Herald Sun, depending on whether you...
Melbourne or Sydney Centric, sorry for Brisbane, I think it's Courier Mail. If you're not willing to see your decisions, the outcomes and the consequences on the front page or in discussed in the pub, then don't do it. It's a common sense guardrail. you don't want to say to junior employees, hey, just go for it. And there are no checks and balances. You need those guardrails there. Business needs the guardrails. Government needs the guardrails.
And society needs guardrails. I'm sure we'll get into regulation at some stage, but think of regulation as being those social level guardrails. And we need to think carefully about those because when you're starting to talk about particularly again, AI and human services, and I keep coming back to that because I just see so many benefits potentially coming out there. You are, of course, talking about lives and in health.
literally about lives. So you want to make sure that the guardrails and the protections are there. And I've heard you talk previously about a great example of the benefits of AI for healthcare workers and what that means in terms of hours per week. you want to share that? Yeah. Because I think that's such a powerful story. Yeah. And it goes back to that. Yeah, it goes back to that. What are the boring bits of your job? And we're already seeing AI really in health and education start to just make it easier to do the boring bits of the job. So earlier this year, you mentioned a few Productivity Commission reports. We also put out two that were focused on health, one on the digital economy and health, and the other mentioning or looking at health productivity. And I'll mention both of them here. So both up on our website. We'll put the links in the show notes. We'll put them either or easy to search and I can use Google, guess. Google is now just an adjective meaning search. You can Google for a Productivity Commission Health Report. We looked at the role of broader technology, but AI in particular. And when you start looking at the benefits of AI in health, there are some frontline benefits. So image recognition, radiology, already been trials to show how the AI and the human together massively improve the diagnostic side of radiology, being able to pick up from scans areas of concern, for example. Fantastic. That means more people will get diagnosed. That means more people will be healthy and well. But on the other side, health has an awful lot of really, really, really boring paperwork. And our health professionals spend a lot of time on that. And we looked at some of the advances that are already happened, some of the predictions going forward. And you could get a saving of around 11 hours per week for each health worker. So this is across the board as an average, but think of it as each health worker saving about 11 hours per week of the drudge time. So they get to spend more time doing the things that they love. They will see more patients, more consumers, more clients, depending on which part of the health system you're in, you do use a different word, but they will see more people that need their help.
And that makes their bunch of better events. And the thing that I heard you say about it previously as well is that's what we actually train them to do. We didn't train them to fill in forms. We train them to use their expertise and their critical thinking in helping a human directly, not to sit in an office and fill out paperwork. And so many parts of our health system we've got, we sort of talked today about labour shortages. Wow. Imagine 11 hours per week on every health worker.Suddenly you're making a huge dent in that labor shortage, which is fantastic. If you ever saw it, so that was one paper. This is love from the guy to have a look at it. The other one is also coming back to that productivity point in services. It's really hard to measure productivity in services because you measure coffees per barista or, you know, how do you actually measure it? Because quality is so important. And we looked at the health system. we went back and we used the data that we could. And it's a fairly limited data set. according to the official statistics, health productivity is pretty much unchanged for like 30 years. And
You sort of think, that's just got to be wrong. You know, if I got sick and someone said, hey, we can pop you in a time machine, take you back 30 years, would you prefer that? you're saying, don't be silly question. Obviously I prefer today. So something's not being measured. So we were able to look at that decade from about 2010 through to about 2019. And for the conditions where we've got data, we were able to look at productivity explicitly taking into account quality of life outcomes. And the number suddenly goes from 0 % per year productivity growth to 3 % per year. Now that may sound small, but 3 % in productivity terms is actually huge and way bigger than the measured productivity across Australia or any of the other developed countries over that same period. So what AI is going to do is it's going to improve these services. It's going to free up labor. It's going to...improve quality, it's going to mean more people and health are well. We may still have trouble measuring those gains, but each of us is going to look and sort of say, I don't want to jump into that time machine, guys. We'll know the gains are there. So what I'm hearing, because, you know, working in organisations for so long, what I've noticed is leaders often grapple with being able to measure the productivity of their individual staff member. which is.reason why we're seeing quite a few organisations demand that people come back into the office because they don't really have that clarity of productivity measurement over what are they doing when they're at home. And so what I'm hearing from you here is that when you have a health worker, in terms of their productivity, we don't really want to measure the number of forms they fill in because that's not actually productive time. Instead, we want to measure more of the activities that get to the bigger picture outcome like that I can't remember what you said now, but that life the quality improvement so there is actually quality adjusted life years how big is the improvement in just Let's say somebody who's got cancer early stages gets picked up by the radiologist with AI together. So the current manual system is double check system, but they've shown that if you add AI and single human check, you raise the detection, the efficacy of by from memory and sorry if this is wrong, but from going from about 80 to 95 % are getting picked up, which is huge. And so that means a person who
has early stage cancer can get detected earlier, they can be treated earlier, the treatment is much more likely to be successful at that early stage. Rather than it being picked up, I mean it will be picked up at some stage down the track, but obviously at that stage the treatment's much worse, so their quality of life is much worse and their life expectancy is much lower. So that's all tried to be captured by this QALY measure. How do you say that not just your length of life, but also the quality of your life has been...
And that's where we're getting these big gains in healthcare. Yeah. Okay. Good. It's also having an impact in terms of less people are suing the diagnostic companies because they didn't catch it initially. Yeah. So I've seen some articles about that, which obviously is better in many respects for a whole range of reasons. And again, it comes back at the business level. It means that if diagnostic companies and more generally, if businesses aren't using the AI tools that improve the outcomes for their customers, they're going to be dinosaurs and they're going to lose their customers. Yeah. And I think another really important part about what you're talking about is in the colleagues is something that we advocate for in organisations, which is that people experience and We always say the technology is amazing. We're advocates for it in the right ways in organisations to help people feel like they're doing their best work, not just the mundane work, but the work that they're really excited about and get up in the morning and go, woof, I'm going to clean my inbox today. No, not that way. But we talk about
the leader's responsibility and opportunity in curating that experience in the workplace and how important that is for all the reasons around culture and none of which, the least is the customer experience. Because we all know if we go somewhere in our services-based economy where you have a great customer experience, you're more likely to go back. And if that person's not feeling like they're supported in their workplace to give the best of themselves and to have the opportunities to further their career their skills and their career and provide a great service, well then they're not going to give their best to the customer experience. So in terms of the question here is really around what are your thoughts on how over time the technologies in the workplace have shaped people's experience in the past and what that looks like moving forward and how that in turn affects performance and productivity and how that scales up in terms of productivity more broadly.
And what's your role of technology adoption in that? And I think we've danced around that a little bit, but yeah, just in terms of that experience side. That longer term, the historic experience, yeah. And the easy way to put it is that we are all just so much better off as customers, as consumers, as people trying to just enjoy life. And we're also way better off as workers due to the long term march of tech.
improving jobs, improving opportunities. And you can see that, you know, go back. Sometimes in developing countries, say, people leave the land and end up in the city in these terrible jobs in factories. When you go out and look at it, there's only one thing worse than having one of those terrible jobs in the factories, and that's being an impoverished farmer burning cow dung at night to try and get some warmth and some light in your often bark or...corrugated tin iron houses that can be caught. you know, and the evidence also shows they don't stay long in those jobs. actually move on into better jobs pretty quickly in most developed countries. In some ways, they're seeing in the narrow what we've seen in developed countries over a period of hundreds of years. So think back to, you know, the Charles Dickens satanic mills of industry. I mean, they were real horrible places to work and life expectancy dropped for people who went and worked there. We don't have them today. Why don't we have them today? Well, basically technology has changed and we mentioned electrification before, which was one of those steps. More recent one is if you go back and look at films or read books around the 1930s, if you've ever seen On the Waterfront, for example, you know, a classic film, and I sort of think kids today must look at it and say, what the...hell is going on? Why are these guys standing outside the gate trying to get into this job and they're manually loading holds of ships. The container was invented essentially around the 1950s, had been some early things. First container in Australia, first container point was actually in Freo, 1969. So you've got a period of getting up towards 20 years. Ports today, another PC route. port you can have a look at if you like is one on the maritime industry where we look at container ports. They are amazingly safe. People literally used to regularly die on ports trying to fill the holds of ships or getting stuff out of a hold of ships. There are still workplace injuries and it be great to get rid of them but you know ports today are amazingly safe compared to what they used to be. The costs of shipping have gone down massively and that's all due to a technological innovation. Now there's a couple of lessons to come out of that. One is that technology and AI is part of that ongoing technological revolution. Some people talk about the fourth or fifth industrial revolution, I sort of lose track. I say, no, in terms of human history, we're still in the one industrial revolution. It started around 1750, and we've just been going ever since. So AI is going to improve the outcomes for workers, going to make work more interesting.
It's going to lead to better outcomes for consumers. But, and the two buts that we can get from history. The first but is that it takes time. So, you know, I was asked a question a few weeks ago, sort of, we haven't seen the productivity improvements due to AI and transformer models, voice recognition type models.
18 months guys! To blip. will say computers took 20 years. 10 years after the PC had essentially been introduced to Bob Solow, who, Nobel Prize winning economist, said, you can see the computer revolution, the personal computer revolution everywhere except in the productivity statistics. It takes time for these to go through. It requires co-investment. So why did it take 20 years? from essentially the invention of situating where you say, actually if we have a box of standard size that we can put on a truck and then put on a ship and take off the ship and buy a crane and put it on a truck and then load and unload it in a warehouse somewhere safely, isn't that a better way of doing things? Well, yeah, it was sort of obvious, but think of the whole range of standards that you have to change in there because first you needed a standard box.
And containers had to be a standard size so you could build ships that would take those containers efficiently. And then you needed to completely change how your port system operated from people manually loading ships to cranes taking containers on and off. You need to store the containers somewhere and so you needed container parks. It completely changed logistics. But the gains, I mean people talk of globalization, globalization was driven by the container. never, we never would have happened. The sort of changes that we've seen in East Asia just could not have occurred with those import and export oriented economies without containerization. But it took 20 years because you had a whole bunch of changes. The third lesson, and again, it comes back to, so I guess, finish that one. We are going to have to have a whole range of changes in our economy around AI to get the benefits out of it.
You can do at your individual business level, but there also needs to be things done at the aggregate level between businesses. And that's going to be really important. A good example of something done well was the ATO's single touch payroll, which was essentially a platform designed to say, to allow interoperability and to say, you know, build on us. We're a facilitator. We're not just
making life easier for you on your tax returns for payroll tax once a month or once a quarter, where providing a platform for business services to build off. That's the best practice we unfortunately often see, worst practice where even at a business level they will take a technology but isn't used elsewhere. other point, sorry, other point is don't expect everybody to be better off. There will be some jobs lost.
There will be some people who say, I used to do this task. I used to work in a typing pool. I'm old enough to remember typing pools. There were literally millions of, at that time, women working in typing pools around the world. Personal computers made them redundant over a period of 10, 15 years. Female workforce participation rose at the same time. Almost all of those women found other jobs and found better jobs because technology creates jobs as well as making some tasks redundant. We're going to see that with AI. And there will also be other vested interests that say, we don't want things to change. so you've got, you've got sometimes called bootleggers and Baptists. So you've got people who saying, we have to protect the jobs because those people are going to be unemployed.
Really agree. don't want to protect the jobs though. You want to make sure there's a safety net and the help to find new jobs and better jobs. pathways, the skills pathways. Are there. But you're also going to have the bootleggers who are sitting there sort of saying, damn, this is going to destroy my business. This is going to make life hard for me. How can I stop this change or get government to pass rules against this change? you know, classic example, I think it was in England blacksmiths and people associated with horses didn't like cars. What you do? Well, you get the government to pass a rule to say, yeah, cars are intrinsically dangerous. You need somebody walking in front of a motor vehicle, ringing a bell to tell people that the motor vehicle is coming so that there aren't any injuries. And you sort of think, OK, that's just destroyed 95 % of the benefits of this motor vehicle. But if you're
But Blacksmith in 1890s, 1900 England and you see your life's work disappearing, you're going to want to try and stop progress too. Yeah. And we have to really grapple with the people who will be disadvantaged. And also there are going to be so many new jobs created that we can't even think about now. Like even jobs that we wouldn't have had about five or six years ago that are prevalent.
today, there will be more and more. And it's our job as a society to, as you say, have the safety net in place for people and have the opportunities. Because people want to do good work. They want their skills to evolve. The number one thing that motivates people is progress. And that's their own personal progress as much as it is progress in the day to day or in that task that they're doing. So it's on us to have those conversations as a society.
We advocate for leaders to do that in organisations to think about really advocating for uplift of skills versus just complete reliance on dautomation. Like you can have efficiencies from the technology, absolutely, and look at your workforce. You've got people in call centres who know the most about your customers. They'll know the most about your products. They are really important skills and knowledge in your business that you can use in other areas.
For new products and services, don't just look at, okay, well, I can go from a call center of 200 people and down to 50 because we've got conversational AI. It's like, maybe you can get some of those efficiencies. What other opportunities can you create in your business? And if you do have to help people leave your organization, you can do that really respectfully. And there is some really good opportunities and ways of thinking about that and looking at, well,
You know, we've talked about it before in another podcast episode, because what leaders often lose sight of is the people that remain in your business are watching the experience of the people that are leaving and going, was that respectful? Was that a good experience? Or was that you're here today, you're gone tomorrow, and we don't care about you as a human. And it damages the culture that you've got in your organisation if people can see that people have been treated really poorly. And that's going to cost you money.
Ultimately. And I think it's also, I mean, this is where the historic slow rate of adoption of technological revolutions works, in a sense, our favour as a society. We do have a tendency to sort of say, back in 1960s, they did it this way. Now we do it this way. And so those jobs have disappeared. But it doesn't happen overnight. And business leaders need to think about, well first I guess they need to recognise that the strong evidence is that at the stage we are in the development of AI, AI and people together is how you get the magic. So it's not like the automated chatbot is going to displace the call centre yet. Maybe we'll get voice recognition and it's gone through the roof due to the large language models.
So it's way better than it used to be, I no longer shout at my telephone quite as much. But it's still got a way to go and customers still like talking to people and the studies have been done to show that the AI and the call center person together, put them together as a team and that's where you get the big productivity improvements.
Again, I'd have to go back to research that's been done out of Stanford University, but I think you're looking like it's 17 % increase in the labor productivity of your core center worker by having the AI assisting them in meeting the customer's needs. And the customer gets better service. So that's a win-win for everybody. And yes, over time, as the AI gets better, you may have fewer people and more of the technology.
but it's not going to happen immediately. if you're a business leader and you're saying, right, revolution, I'm getting rid of this whole group of people, you're probably wrong. And even very, very savvy tech people get this wrong. So one of the founders of the internet back in 2016 said, we should stop training radiologists because there won't be any jobs for them in a few years. We know, as we've already discussed, it's AI and radiologists together that creates some magic. So yeah, we need radiologists and guess what? Their jobs are probably a lot better now because they've got rid of the boring stuff because the AI is helping them. And if I can connect back to something that you said earlier, it sounds like for leaders to be able to get the productivity out of this new technology, takes right now what they can do is some analysis and some really thoughtful decision making around what are my individual staff members doing?
What's core in their role that's producing the end goal and quality of life and outcome that we're trying to drive and what are all the things that are taking up time and space that we could look at reducing or automating. So it feels like that analysis and thoughtful decision making could happen now before they get to anything else so that they do start to do it in a more strategic way that's contributing.
It's not just the business leader saying, will think about these things. It's the business leaders and the team members working together because the people who can really give you the answers, the people sitting three levels up the hierarchy who aren't quite sure what the person actually is doing down there. So it's everyone's role. It's everyone's role. And you will get the best ideas.
from the individuals who are working in those roles, as long as they see firstly that their ideas are welcome, that it's a respectful, it's a trusted environment where they can say, hey, what about this? And it's not simply, we don't do things that way, you know, go back to your desk. So it's got to be that internal environment that invites questioning about how things are done and allows things to be improved. And it's got to be the case that
workers themselves, they don't see this as a job replacement exercise because if they do, the last thing they're going to think up is good ways to replace themselves. Self-sabotage, yeah. no, we couldn't possibly automate that part of the process. Yeah, and you know, there is such great opportunity here to really land, you know, whether you call it human in the loop or augmented human intelligence, the reality is both
in our work, how we work, live and play, to quote Kylie Walker from ATSI, is changing. And more and more it's interconnected and codependent. And we can't think about them as separate things. And yes, AI will increasingly over decades play more of a role in some of the really mundane and just boring tasks that humans don't want to do.
We have to make sense of that in small incremental steps each day. And at the same time, we have to create hope and curiosity and interest for people in what their opportunities are in their career and how that evolved and be part of that conversation versus just saying, we're going to bring all this technology in, you're going to, that comes in that door and you're at that door. It's like, it doesn't work that way. And it's also, it's respect across the spectrum of your workforce.
recognizing that the revolution doesn't have to happen tomorrow, that the co-investments that we require, the change in systems, change in ways of doing things across the industry, across the country and across the world for AI, that's all going to take time. So there will be people in your organization who say, God, I just don't want engage with this. I'm five years from retirement.
Yeah, they should feel it respected as well. shouldn't be a, so okay, you're a dinosaur, you're out of here sort of approach. You you need to tap your workforce for ideas, but you also need to show your workforce, hey, this is a way to make your job more interesting. And yes, I know that it's been done a different way for most of your career.
And yes, I know that it's hard to change, but we're going to support you in changing the way that you do your role so that you can really have a great last end of your career rather than sitting in the corner grizzly. And I can't remember this study. I'll have to come back to it.
We make a lot of assumptions about older generations not necessarily willing to embrace the technology when in actual fact they quite often are. Because they know that there's better. They're older and wiser for one of a better way of saying it, but they can see that there's better things to spend their time on. And maybe that's the thing that's driving them to want to embrace it. It's like the clock's ticking, so my legacy and what I can do here can be supercharged by doing it better. We're seeing that more and we're seeing also in history, now is a time when we've got four and five generations at one time, working in the workforce, which is something we have to encourage leaders to be thinking about deliberately for how do you be inclusive of that in your workplace? Because you've got some young people coming in who expect a lot more automation than where you're possibly at.
based on how they live their life outside of work. And so you've got younger generations with high expectations and then you've got lots of experience and knowledge over the subsequent generations where they can add value and they can add nuance to that. Think about that from the perspective of the human experience and the experience of those different generations in the workforce really add the flavour and the diversity of thought. And it's up to leaders to be able to tap into that, to your point give them the opportunity to participate in ways that they can feel that they've got access to it and they've got access to learn about it. We bang on a lot at the moment about AI literacy and giving people a lot of access to what the technology is. Yeah, it's really helping all of your workforce feel empowered to embrace the technology, embrace the change and that their views will be valued. And yeah,
We've got to make sure that we don't fall into stereotypes around groups of workers and you mentioned age and I'm sort of getting towards the end of that edge of the spectrum but when people sort of say, Stephen, how come you're involved in technology? I sort of say, well, one of my neighbours across the road is an electrical engineer by training who was working in two, a couple of years ago, think, 89 when he finally finished his last contract.
He's the one who introduced me to AI research tools. He comes running across the road one day saying, have you tried this product? It's unbelievable. so, yeah, and it's tapping into the diversity and the enthusiasm of your workforce and a good business leader will.
tap into that and be able to bring out the enthusiasm because the enthusiasm is always going to be there. just have to, know, different people will have different ways of bringing it out and it's being able to manage that. That's what management's about, it's people. Yeah, it's leadership is that as a leader you're curating the experience of your people through how you support them to feel like they're included, but also to experiment with different technologies, to ask for help and all of those things. And also owning the work they do. think we need to make sure that accountability is in there as well. We always ask our guests when thinking about humans and AI in the workplace, what's your wow? What's your wins, your opportunities and your watch out? OK, so the wins will be there for the business leaders who are able to have that inclusive conversation about technology, about how they can get the best benefits out of that technology within their business, and will also help facilitate the broader social discussion. Because at the moment, the opposite of wins obviously are loss, the risks and the potential loss is that a lot of the debate around AI is based around the machines are coming for our jobs, we're all going to be, you know and quite ludicrous scenarios. the AI is going to blow us all up. OK, so you put the AI in charge of nuclear codes and didn't think it. Most of your listeners will be too young to remember the film Doctor Strangelove. Go and watch Doctor Strangelove or watch the clip about the Doomsday device on Doctor Strangelove online. It's a comedy, but that's driving debate. So the win is that internal organization, but also get involved in the external race. Get out there and say, hey, actually AI has really helped our organization. Our customers are better off. We've got the safeguards in place. Our workers are happier. That's where have a win. And that really comes back to the opportunity because that is the opportunity around AI, that it's a win-win technology for business, for customers, for workers. And that's the great thing about a general purpose technology. So wins, opportunities and? Watch out. Watch out. Watch out. OK. The biggest watch out is around regulation. And when you have a debate that's often dominated by overly cautious fears, if I can put it that way, the risk is for politicians to take action.
to be seen to be doing something and you end up with bill thought out regulation. It's almost impossible. You know, I'm a former regulator. I was at the ACCC for five years as a commissioner there. And I say, well, we have consumer protection regulation. I was involved in it. We have competition regulation. I was involved in that. Other regulations around privacy, copyright, IP. There's a whole bunch of regulation out there. And the challenge I tend to give to people is tell me the harm that's supposed to be going to be created by AI that isn't already covered by existing rules or and it's got to be actually more than that. It's going to be isn't even envisaged by existing rules because rules are just and we see rules adjust over time. So the starting point is we probably have most of the regulation we need there. There will be gaps and we'll find them over time. We'll fill them and hopefully on a technologically neutral basis because there's nothing worse.
than passing a bunch of legislation that 10 years later is completely obsolete because the technology's moved on. But that's the risk there. That's a watch out. Watch out for regulations that kill the benefits before they can be created. We don't want that person walking, ringing a bell in front of the AI. That's right, yes. That's a classic.
Look, we could talk all day. There's so many questions that I still have, would love to ask. But in the interest of time, it's been so good. Thank you so much for your insights. I can already, I'm already getting excited thinking about, you know, how we can share leadership insights from a productivity commissioner in the topic of AI, because you've shared some really important goals there and one which I
I'd love for us to take into the workplace, is the quality adjusted life years. How can we really embrace that in the workplace? think there's a really good opportunity for that to be part of our language moving forward because work is such a huge part of our day and our lives. And so thank you so much. Great. Thank you. for that. Thank you.
Humans and AI in the Workplace is brought to you by AI Adaptive. Thank you so much for listening today. You can help us continue to supercharge workplaces with AI by subscribing and sharing this podcast and joining us on LinkedIn.
