How to redesign work so AI creates real value for your organisation

About the episode

What sets your business apart when your competitors all have access to the same AI capabilities?

AI is the biggest technological shift of our generation, and it could completely revolutionise your business. But simply bolting AI tools onto your pre-existing workflows won't give you a competitive edge. So how can you redesign the way you work, with AI in mind?

UNSW Business School Dean Professor Paul Andon says: "The trap is rushing straight into application. We need to think first, act next."

If the future of work is AI-driven, how will you adapt?

Interested in hearing more about AI and work? Listen to our episode with Scientia Professor Toby Walsh, Chief Scientist at the UNSW AI Institute  

The Business Of podcast is brought to you by UNSW Business School, produced with Deadset Studios. It’s hosted by Professor of Practice Fahim Khondaker.

Want to know more? 

Want to follow the latest research and news from UNSW Business School and AGSM @ UNSW Business School? Subscribe to our industry stories at BusinessThink and follow UNSW Business School and AGSM on LinkedIn. Or listen to more episodes of The Business Of and subscribe to our UNSW Business School and AGSM channels on YouTube.

Edited transcript

Professor Paul Andon  00:00
We don't have a clear path forward with artificial intelligence. You need to think strategically about the direction of travel, without knowing for sure what the destination will be.

Professor Fahim Khondaker  00:11
Right now, AI adoption in business is on a wide spectrum. On one end, it's a productivity hack: drafting emails, summarising meetings, and cleaning up a dataset. Useful, but it's not where the advantage is. And if that's where your business stops, you're not behind in AI, you're just slightly faster at what you're already doing. On the other end of the spectrum, the businesses pulling ahead have stopped asking, "How do I use AI?" And they've started asking a much harder question: "What can my business do now that it genuinely could not do before?" And, "How can I transform my business using AI?" It's not an automatic shift, and it's not without risk either. The more AI does, the more a business needs people who can still think without it. This is The Business Of: AI Explained. I'm Fahim Khondaker, Partner in Data and AI at BDO, and Professor of Practice at UNSW. Joining me is Professor Paul Andon, Dean of UNSW Business School. He's spearheading UNSW's integration of AI into business education, preparing leaders for the future of work. Welcome, Paul. Great to have you here today.

Professor Paul Andon  01:22
Thanks, Fahim. Great to be here.

Professor Fahim Khondaker  01:24
If you walk into any office today, you're pretty sure everyone has Claude, Copilot, Gemini, or something similar open, using it to do their daily tasks as they go about their day. You've spoken about this middle ground, or uncomfortable middle ground, where things have changed slightly. Can you tell us a little more about what you think has changed recently?

Professor Paul Andon  01:43
Yeah, sure. So AI has only been around for about three or so years. I think a lot's changed in that time, and we're seeing a bit of an evolution in how people use it. So, of course, to start off with, everyone was just playing around with it, experimenting, and trying things out. Firstly, in a very playful way. I think very quickly, then it turned to: "How can I use this in my work? And so, that's sort of translated into how it can help people to write? How can it help people create the documents they need for their everyday activities? I think that as AI has evolved, people have started to look at how they can use it for more advanced tasks. It's very good at putting together slide decks now. It's very good at helping with qualitative analysis of a lot of information that they may have at their fingertips, and so on. I think we're at a point now where organisations are moving away from individualised productivity uses, and you're starting to see the conversation turn to, "Well, how do we generate organisational value out of this tool?" Many organisations have allowed their staff to free play, then they've started to introduce enterprise licences, and, of course, you have the added impact of not just the licence but also token consumption. And so now we're at a point where organisations are starting to ask themselves hard questions: "What do we do now to extract value out of this new technology?" And I think that's the uncomfortable middle ground, in a sense, that people are in because they still haven't probably quite realised the full potential of individual productivity. But they don't always know where to take it, in terms of the organisational value extraction. Certain organisations are certainly leading in that space. You look at your CommBanks, your Telstras, your Fortescues. They're investing heavily, but not all organisations have that capacity to invest strongly right from the outset.

Professor Fahim Khondaker  03:33
What are the big challenges that organisations are facing in adopting AI systems?

Professor Paul Andon  03:37
So I think the biggest challenge is uncertainty, and I think people really want to understand where they need to go with A. But of course, the answers aren't there because this is a major technological change, as we see with other major general-purpose technological transformations. You know, if you think back to the information age of the early 2000s. There was uncertainty about where that would go, but eventually things changed. And I think the key thing now is we don't know what that equivalent transformation is going to be, in precise terms. So the uncertainty means that organisations have an important choice. Do they decide to lead now? Or do they wait for clarity to come? Of course, there are risks and benefits on both sides. If you lead, you may make the right strategic bet, and you're an industry leader. If you wait, then you may be left behind. So, we have to be willing, as organisations, to make mistakes, learn from those experiences, and carry on. So I think that's an important role for the CEO. The CEO, really, I know that's a very sort of cliched sort of thing to say, but the CEO really needs to set the tone. What sort of organisation does that entity want to be in an AI-enabled world, and how quickly and how much of a leadership position do they want to take?

Professor Fahim Khondaker  04:55
So that's a really big, big-picture challenge.

Professor Paul Andon  04:58
Yeah.

Professor Fahim Khondaker  04:58
What are some incremental challenges in terms of implementation, and what are you seeing are common challenges that most organisations face? Let's say I've got a boss who's clear that we've got a clear vision. We're going to invest, we're going to spend money on this. What's next?

Professor Paul Andon  05:12
I guess then it translates down into what does that mean for individual workers? What does it mean for organisational activities? How do you think through the transformation journey that that, you know, workforce and workflows need to go through? And I'm reminded, just thinking about it now, of an analogy that was in an op-ed in the AFR, where they talked about the invention of elevators. And you know, to start with, architects just bolted on elevators to existing buildings, to make it easier for people to get up and down your three or four-story, you know, apartments, etc. The full value of that new technology only came to the fore once architects started to rethink the design of buildings, the urban architecture that would go around it, and the way in which people could live and work, now that we had this new technology, which means we could build in much greater density and at much greater scale. So, there's that transformative piece that I think we need to consider in organisations, given that we've been spending time bolting this technology onto existing workflows. How do we actually rethink the way in which things are done? And it's probably easier for organisations that are new and starting out. And starting to sort of build their capacity and scale. For existing, well-established organisations with established cultures and ways of working, that's a much harder thing to move.

Professor Fahim Khondaker  06:41
And then what about on the technical side? Do you see any challenges in adopting the technology itself? Have you come across anything like that?

Professor Paul Andon  06:48
From the technical side of things, I'm not a technical expert so far as AI is concerned. But I think people understand what it can do, what its limits are, but also what the opportunities are. So I think there's a level of organisational reskilling that needs to be done. That, you know, many organisations either haven't had the capacity, or people haven't had the time, or there's just not the resourcing required to do that wholesale risk, so people fully understand the capabilities of AI. Probably the other aspect, from a technical point of view, is just making sure that organisations feel comfortable with privacy, cybersecurity, and all the different sorts of governance-oriented considerations we need to think about. And again, it's not always entirely clear because the AI is, in a sense, a bit of a black box.

Professor Fahim Khondaker  07:39
I really want to unpack public sentiment around AI, the challenge it poses to organisational leaders, and how you manage it. Can you give us your perspective? Obviously, you're leading at a university, rolling out AI, and it's obviously no secret that everyone knows that university; the sentiment around AI use in universities is a very hot topic.

Professor Paul Andon  07:59
Yes.

Professor Fahim Khondaker  08:00
I'd love to hear from you. Can you give us some insights? What goes through your head as you face these, and you see these on social media and things like that?

Professor Paul Andon  08:07
This is a major change happening in society. That uncertainty creates anxiety. That uncertainty creates concern about what the future might hold for people, the nature of their work, and how it will change what gets done. You know, we talk about things like academic integrity in university settings and students' capacity to learn. I think they're important challenges that universities need to take on and solve. But by the same token, we should take the concerns raised very seriously because, you know, I think about the transformation of organisations across a couple of different themes. We've talked a little bit about the workforce. We've talked a bit about workflows. The thing we've only just started to touch on now is what I clumsily sort of label as work-life. It's not just a transformation of a very mechanical sort of thing when we're talking about AI. It's something that fundamentally challenges human agency and human intelligence, so we need to understand how to bring the best of both worlds together without necessarily downplaying one side of that balance or the other.

Professor Fahim Khondaker  09:21
I've spoken to you before, and outside of this forum, about navigating this uncertainty, and you offered this lovely analogy: we need a compass right now, rather than a map, because we don't know where we're going and just need direction, right? Can you expand on that a little bit more, in the context of what we've spoken about today?

Professor Paul Andon  09:39
So we don't have a clear path forward with artificial intelligence, whether it's in the university sector, whether it's in telecommunications, pick any industry. The future of that industry in an AI-enabled world is yet to be told. So, with that in mind, you know, you need to think through strategically what the kind of direction of travel is, without knowing for sure what the destination will be. So, in my role as a leader in the Business School at UNSW, before I became Dean, I started a conversation with the leadership group about our direction of travel on the education side of our work. And so, we came up with a very simple direction statement. So, that kind of gives us a compass for the direction of travel and what we need to do, but it doesn't specify the outcome. Other than, we still want to be a leading business school in Australia that provides high quality business education, and we know that that's going to be different in an AI-enabled world. We need to do the hard work to understand what that looks like in the coming months and years.

Professor Fahim Khondaker  10:45
Paul, we've seen a significant shift in how people use AI, as you said. Let's get a little bit functional. Let's start with, if you're in a particular part of a business, what are the ways you can get a competitive advantage for yourself, and for your organisation, using AI?

Professor Paul Andon  11:00
So I think the trap is to rush straight into the application. I think before you start thinking about the application, people's minds need to turn to your overall value proposition. And is that changing in an AI-enabled world? So if you think about all the seats at the table, and you know your marketing function, your finance function, IT function, human resources, people and culture, all the different functions of an organisation. What is your value proposition around that table? And how does that change as a result of artificial intelligence? Now, I can probably dive a little bit deeper into the accounting profession because that's sort of my home discipline, so to speak. You know, I think there's a real challenge for the profession now in defining its proposition going forward. There's a lot of talk about automation, and about how you know the basic work of accountants isn't required anymore. Yet, what we see in the marketplace is a shortage of accountants. So, I guess the question you need to ask is why. And I think maybe what we need to understand in that space is that the value proposition we think accountants provide is actually wrong. It's not about maintaining accurate, complete records and ensuring transactions go through effectively. And that's an important function of accounting and finance professionals. But their real value-add is how they help organisations understand and run their businesses successfully. Because they provide the intelligence and information that help the entity understand what's going on in an organisation, which, in turn, allows you to make very informed and intelligent decisions. So I take that as an example of every seat around the table. What's your value proposition? How is AI going to change your delivery of that value proposition, or change the proposition itself? And then, how do you transform to realise that?

Professor Fahim Khondaker  12:52
And is there a risk that they could get it wrong? And what does that look like?

Professor Paul Andon  12:58
So the risks are, you know, for example, you may over-index, and we might decide that AI can do a whole bunch of things that actually turn out that it can't do as well as we thought. And so, then we might look at reducing our headcount and going a bit too far on that front. So there's that sort of risk in striking the right balance between how much you utilise AI and how much you still value and require human expertise. I think there's a potential risk also, in getting a bit too caught up in the hype and just following the hype that that goes with, you know, how a particular vendor might say that, you know, AI will transform the way you do things. I think there are no shortcuts here. Organisations do need to think through, and professions do need to think through for themselves very carefully, you know, what the value of AI is, can, and should be for them. And to sort of not get too caught up in the glossy, vaporware-type presentations that many big tech and other organisations can potentially use to excite people.

Professor Fahim Khondaker  14:00
Let's dive into that a little bit more then. How do we adopt AI in a world where there's a mix of humans and AI agents working together, or just AI embedded into the function, as you were talking about? What are some of the practical things we should be doing?

Professor Paul Andon  14:15
I think this is one of those things where the more things change, the more they stay the same. So, it's good old-fashioned value chain analysis, strategic thinking, workforce planning, upskilling, and capability development. These are all very stock-standard managerial practices, and I think, to some degree, we kind of forget the importance of those things when we get, you know, excited about new technologies. You know, I think it's a case of sitting down and thinking carefully about where things are best automated. Where things can be done, where the expertise of an individual can be best augmented. You may use AI to help you think through the possibilities to start with. And I think the hardest thing then is: what new tasks would emerge as a result of AI changing the workflow of a finance function? So, that's the way I would think about it: value chain analysis, strategic thinking, workforce planning, and upskilling. And then from a sort of managerial framework point of view, just going with a good old-fashioned value chain analysis, and thinking deliberately and carefully about where those things need to change at different points, in a workflow.

Professor Fahim Khondaker  15:28
When we talk about value chains, there are obviously steps to the value chain, and I love this example that we've come across about coffee and a coffee business. Coffee is a bean. You grow coffee, harvest it, roast it, pack it, deliver it to a cafe, and then you and I order it. So that would be a standard coffee value chain. And the way I talk to people is that you can use AI in each of those steps, or you can use it across the lot. And is that, kind of, what you were thinking when you were talking about value chains?

Professor Paul Andon  15:55
Yeah. So I think it is, and I think there are two things out of that. One, you kept it very simple. So I think simplicity is important too. Then I think it is just, as you described it. You come up with a general, staged process, and you think through not just those individual processes but the linkages and synergies between them, and how AI can enable change in ways we perhaps couldn't before. So I think the way you described it is actually absolutely right.

Professor Fahim Khondaker  16:24
Quite often, people talk about AI in the context of speeding things up. It's a very efficiency-led conversation. But I know you have a view that that's only part of the picture. There's such a quality argument, where AI can be used to improve the quality of what you do. Can you talk to us a little bit more about that one?

Professor Paul Andon  16:43
Yeah, sure. Well, it's not just me. It's one of the three national AI priorities, and it's taking advantage of the opportunity. So I think, you know, we often think about keeping people safe, and I think that's really important. But on the benefit side, as you say, because I think we kind of bring a tech mindset to all this, and tech in the past has always been about efficiency and automation. Whereas this technological change is something different. It brings, and I use this term loosely, intelligence to the table. So people can actually utilise AI to support, augment, and assist with creative tasks, and to carry out work, and that can be done in ways that increase not just the quantity of work you do, but actually the quality. So I think that quality aspect really needs to be borne in mind, and I think it's important not to think that the AI does all the work as a result. So the quality doesn't come from the technology; it's the way in which you populate it with context, with your own knowledge, your own awareness of what the challenges and opportunities are in the particular application that you're applying it to. So, the opportunities is as much the human, or the expert thinking about the way in which to best use artificial intelligence, and to provide it with the context and the richness of information and the nuances that you have in your head, so that when you ask ChatGPT or Claude to do something, it has as much information as you do, to help you carry out the task.

Professor Fahim Khondaker  18:22
Let's talk a little bit about thinking ahead, a little bit more about where we are actually going. If all goes to plan, with the way AI is being sold to us, we are going to have AI agents and AI colleagues, effectively, for lack of a better word. And I hate humanising AI, but let's just say we will have AI colleagues. How do you prepare for a world like that? That's such a big change.

Professor Paul Andon  18:44
To be honest, I don't know. And I think this is part of the uncertainty, and it's something that we need to work through over time, as we see the use and the agency of AI evolve. So, how do you prepare for a world where you don't quite expect what AI is going to do when you work with when you work with it? And I think the clear answer is, we won't know until we work through with it, and understand where the, you know, the sort of unintended consequences are. How do we bring AI back to try to mitigate that? How do we put governance in place to help prevent that as much as possible? I think that's all we can do for the time being. But how exactly will we work with agents? It's not clear. But then we have lots of things like, you know, managers need to prepare for a world in which they're managing people and bots together, and, you know, working on different tasks. And I think they're kind of just visions and ideas at the moment, and it'll take time for organisations to figure out the true reality of all that. And indeed, if it is a reality at all.

Professor Fahim Khondaker  19:49
From your perspective, as someone who's involved in an established organisation, and works with mostly organisations that are established. Your advice to anyone if they were to start something from today, from scratch, a workflow from scratch.

Professor Paul Andon  20:01
Yes.

Professor Fahim Khondaker  20:02
Where would your focus be?

Professor Paul Andon  20:04
My focus is to run with your imagination because you don't have the constraints that an established organisation does. So, of course, you need to be responsible. You need to think about the risks and all that. So, you know, I sort of don't discount the need to be careful and need to have constraints. People have the opportunity to really play because what AI does is probably give you the opportunity to do things that maybe you couldn't do before. So that you can, sort of, build up things without necessarily having to have the kind of setup costs that you would normally have in a pre-AI startup situation.

Professor Fahim Khondaker  20:39
And within an established organisation, is there an opportunity to reimagine the entire function?

Professor Paul Andon  20:45
Absolutely. Absolutely, there's an opportunity to reimagine. And you see this again with, you know, great transformations in business models in history. Is that, you know, you go from Blockbuster, if you think about back to the video days. They had a particular business model. There was a technological change. It wasn't Blockbuster that came up with online streaming. It was Netflix, because it started from a point where they didn't have the same constraints that Blockbuster did in terms of their business model, and what they understood generated income for them. They were free to think about new possibilities, and so I guess that's kind of the analogy to help me, kind of, explain, you know, the opportunity that startups have now.

Professor Fahim Khondaker  21:28
I think for me, you've just triggered. Well, okay, startups can do that, but I have to figure out ways to be innovative in my existing business as well. Would that be the first tip?

Professor Paul Andon  21:37
I think there's a more fundamental principle, and it gets back to, I guess, one of the big concerns at the moment, and that's around cognitive decline. And you know, what's AI going to do to professionals who just keep using AI all the time? So I think the first tip we need to make sure everybody bears in mind is: think first, act next. And I think, thinking first means, before you consult AI, just take some time to think about what it is you want to do, how you want to achieve it, what you know about that particular issue, and how you're going to execute it. It may not necessarily be in an orderly fashion. Indeed, this is how I use AI a lot. I have a lot of ideas and a lot of thoughts about a certain issue, and I'll just dump them into AI via voice recording or chat, as in voice chat. Then I use AI to help me organise it, which brings some clarity to my next iteration of thinking about how I take on a particular task. So I think "think first" would be my first and most important tip. I think that's our, that's our antidote against the concerns for cognitive decline, whether it's in education, whether it's in professional expertise, and so on.

Professor Fahim Khondaker  22:50
So, if I'm the head of a marketing function, a finance function, or some other function in a typical organisation. From today, if I haven't started thinking about adopting AI across my value chains, or I've just been using it for emails and personal use. Where do I start, in your view?

Professor Paul Andon  23:10
Yeah, I think you should start by considering the value AI will bring you. And even before that, the value of your function, or your particular area of work. Think about what it is you're trying to achieve as a business unit in an organisation, and the value that you think you bring to the table of that organisation, and then how AI will help to support that?

Professor Fahim Khondaker  23:33
And any advice on bringing people along on the journey?

Professor Paul Andon  23:36
It's not just a tech rollout. We actually have to build trust. And particularly now, even more so, we have to build trust and confidence because, you know, the temperature around concerns about AI is rising. And that means bringing people along, talking through things, ideating what the possibilities are, thinking through the risks, like you were talking about before, in terms of you know, do we decide to make a play now, or do we wait and see what other similar organisations do, and then follow suit? So if your decision is to wait, I think that's perfectly fine. As long as you've taken the time with your leadership team and done the diligence to determine that that's the right move for your organisation or for your business unit. For others, that decision mightn't fit with their values or ambitions, so they would take a different path. I think, whatever the decision is, as long as it's deliberate, as long as it's thought through, and it has a diligence to it, and brings people along because you are talking to people about that decision, and the issues that go around it. Then I think that's the most important thing.

Professor Fahim Khondaker  24:47
What does AI-enabled work in five years from now look like in Paul Andon's mind?

Professor Paul Andon  24:53
I think there will be new tasks. So we may still have accounting, marketing, and information systems experts. But the nature of their work will fundamentally change. I think we'll see, you know, flashes in the pan. You know, we had prompt engineering, that's probably the shortest-lived profession in history. And we may see other versions of that. But I think fundamentally, we've seen positive transformations of productive activity. Now think back to the, you know, transformation in the information age that created new forms of work. And we had the gig economy. We had the sharing economy; we had, you know, transformations in the way we manage knowledge. And I think, in five years' time, we will see similar sorts of transformations in ways of working, the emergence of new business models. So the flavour of it will be different, but I think the nature of the changes will be similar. So, it'll be exciting to see, and looking forward to coming back in five years' time and having this conversation all over again.

Professor Fahim Khondaker  25:57
So, a general transformation with flashes in the pan, and mostly positive from what it sounds like?

Professor Paul Andon  26:03
I think so, and I think history again suggests that that's the case. It doesn't mean it doesn't come without friction. But, you know, historically, we see greater prosperity, but it doesn't come without significant hard work, and it does require a structural shift. So we need to be mindful of all those aspects. I think we need to be more grounded and realistic, and think about how much we can shape and how much agency we have in shaping our future. And make sure that we are shaping in positive ways.

Professor Fahim Khondaker  26:36
Professor Paul Andon is the Dean of UNSW Business School. If you want to hear more about AI in the workplace, listen to our episode with Chief Scientist at the UNSW AI Institute, Scientia Professor Toby Walsh. You'll find the link in the episode description. The Business Of is brought to you by UNSW Business School and produced with Deadset Studios.

To stay up-to-date with our latest podcasts as well as the latest insights and thought leadership from the Business School, subscribe to BusinessThink.