Exporting intelligence: why Australia’s software edge matters now

Well, good evening. It's a pleasure to be here tonight for the game changers event, the global game changers event, exporting intelligence, Australia's software advantage in a shifting global economy. My name is Alicia Barry and I would like to welcome you all here in the great hall at UTS with me tonight for this event. Uh, and also welcome to those who are joining us virtually as well. It's a pleasure to have you with us tonight. I'm a business and finance host at the ABC and it's quite a treat to be out of the studio tonight and here with you. I'm really looking forward to tonight's conversation. I'd first like to start by acknowledging the Gatagle people of the Aura nation whose land we are on tonight. I would also like to pay respects to the elders past and present and acknowledge them as the traditional custodians of knowledge for this land. And I'd also like to acknowledge the traditional custodians of the various ancestral lands that our online attendees are on at the moment and pay respect to those elders past and present. So it's an interesting time to be an exporter at the moment with Donald Trump's tariff and trade policies spawning protectionism and sending financial and currency markets into a spin. We wake up to headlines daily about the latest development in the US China trade war. So how is this going to affect Australia's software exports? Well, you know, heavyweights like Atlassian, Canva, Wisec Global, even technology 1. But more broadly, as Lucy Ellis has pointed to, digital exports topped $7.5 billion dollars last year, proving that code travels faster than commodities. So, tonight we are asking, where is global demand heading for software? How is AI transforming productivity? And what will it take for Australia to sharpen its edge and scale faster as a software exporter? Well, we have a very esteemed panel to answer those questions for you tonight. Firstly, we have Dr. Terry Roach, the founder and CEO of Capsify, a Sydneybased software company. Now, Terry has a PhD in enterprise architecture and more than 25 years of experience in software and business design and he's obviously taken Australianbuilt software to the world. Welcome, Terry. Next to him is Lucy Ellis, the chief economist at Westpak. She has spent three decades at the Reserve Bank of Australia prior to this role, including as assistant governor and brings deep expertise in trade, productivity, and policy to help business and government navigate shifting economic conditions, which of course is what we're here to talk about tonight. And Professor Asf Gil leads software engineering here at UTS. He's an expert in agile and secure digital systems, working with industry and government on digital resilience and preparing Australia's engineers for an AIdriven future. And then David Masters is next to him. He leads global policy and regulatory affairs at Atlassian. With more than 20 years of experience in tech, public policy, he shapes strategies on trade, AI, privacy, and the settings that help Australian tech companies scale worldwide. Now, the way that this event will work tonight is we're going to engage in some discussion for 45 minutes or so, and then I would really love to bring in your questions. So, please uh send them through. I think we might have the Slido app uh available for you. I think we can bring that up on a slide at some point. But that's the mechanism in which to uh submit questions and I will get to as many of those as I possibly can. But I will take a seat now and we will begin. All right. So Terry, I'd like to start with you. You uh I have heard you describe yourself as an overnight success 11 years in the making. So can you can you tell us how did you start Capsify? What did you do to get to this place now? Sure. Well, I was probably untypical. I was a probably a bit of a late starter as a founder. You don't normally see people my era uh with the with the um audacity to go and think they can um found a business. Um but I I was midway through my career. I kind of came across some problems that I thought looked exciting. I couldn't figure out how to solve them. I decided to go back to uni and late in life did a PhD and that turned into um some research that I managed to commercialize and turn into a business. And so uh for about 12 or 15 years we we gradually built a product on the basis of that that research and we eventually sold the company late last year. Well congratulations. That's wonderful. I imagine Lucy that the global macro environment that exists now is pretty different to what it was for Terry about 13 years ago when he first started uh on his endeavor. Can you give us a bit of a snapshot of what it's like for exporters at the moment given the, you know, global headwinds I just mentioned? Thanks, Alicia. Look, I think it really matters whether you're a an exporter of goods or an exporter of services. And I think that's a really important distinction. It's very hard to tariff a lot of the services that Australia um actually has a growing industry in. And it's not just the $7.5 billion of software licensing exports that's come from nothing 10 years ago, but you know over $8 billion of a category that the ABS has such sort of management, consulting and accounting and auditing. You know, we're a highly educated white collar society. We speak English. So where there are niches, there are opportunities. And I think particularly where you have services that are able to be delivered digitally and remotely, you know, it becomes easier to locate wherever the right skills are or the right environment is. Uh and it, you know, it becomes less relevant where you're buying that software from. Uh and and you see there's also, you know, vibrant software, you know, companies, you know, small and mid-size software companies in Europe as well. And I think people focus so much on you know those you know few big behemoths they don't realize just how much value ad is being made right across the industry. And it only takes I mean when we think about that seven half billion dollars you know you go to the public you know the public uh accounts of you know Canva and and Atassian and Weise you know it only takes a few winners to suddenly have that industry with that kind of export space and you know there was no government plan to create that software export industry and yet it's bigger than barley it's bigger than cotton it's bigger than aluminium. Well, I'd like to pick up on that sort of idea. I mean, Sgill, how is Australia's software industry perceived because you think exports and you think stick some iron ore in a container and whack it on a ship to China? Of course, if we burn them on a plastic, then we can export. But unfortunately, we don't do anymore or uh that's my time when I started uh my career. But I think it's living through the cloud can deliver over the internet. So I think there's a less restriction around there. It could be consumed anywhere. So there are software or I would say that export controls there. Uh but I think still the software could be consumed. Uh just looking at the local Australian industry. So we are currently a $30 billion economy about 1,600 around companies around there plus and we expecting that that would be by 2030 would be about 40 billion uh dollars economy. So there is an appetite is going to be increasing uh there are a lot of digital initiative Lucy mentioned about digitization so that going to actually increase the demand uh for the software so I believe I'm very happy in a way as a software engineer and training people that there would be a lot of demand for software engineers and uh yes there is a concept of AI can enhance productivity and we talk further on that we will get to that David can you give us a bit of a picture of what the regulatory envir environment is like for software companies, developers, those starting out, you know, what's working and what's not. Yeah, I think to Lucy's point, I mean, there wasn't a real plan to to grow the Australian software industry. Sort of happened organically and part of that is because it was a fairly lowly regulated environment. So, I think I think the bar was really low. I think it's becoming higher. I think what you're seeing is a sort of global patchworker regulation. Um, take privacy, cyber security laws. So to operate at a global scale, you have to meet a whole bunch of competing regulations that that are increasingly inconsistent. So I think there's some challenges there. I mean, one of the things we're sort of pushing for is, you know, like take like uh security incident reporting for example. Um you know uh India has 6 hours, Australia is 24 hours, Europe's 72 hours. There's a whole bunch of different regimes where you know you spend a lot of time trying to figure out rather than responding to the incident itself figuring out how quickly you have to report to certain regulators of certain instance depending on the nature of it. So, so I think what we're saying to governments is that actually that kind of patchwork is not helpful. It's actually not going to give you the best outcome because you actually want companies to be dealing with those issues rather than reporting to regulators. So I I think there's um you know there's an emerging um because it's new regulation. It's also as companies have emerged, they haven't built the engines to deal with that regulation. So we're all trying to figure that out on the fly as well. And AI is sort of the new front um of that regulatory regime with the the European AI act and and considerations for other you know state legislation in the US um possibly regulatory response here as well. So it sounds like there's quite a lot of opportunity for Australia to step up and you know fill demand. Lucy Ellis, how can services and software step into that role where the demand is? Well, I think it really comes down to identifying the market and the need and being in a position with the skills in in Australia to actually deliver that. And you know I I would emphasize that you know we're a small economy but we are a highly educated one. Uh I think one of the things that you know distinguishes you is that we have you know you know while there are complaints about the regulatory environment by and large our institutions are reliable. We have a very pragmatic set of sort of you know regulators and you know our sort of you know our political process is very pragmatic and so you know there's a bit less sort of uncertainty around how things will play out in a regulatory sense. I mean we do have a relatively safety focused lowrisk appetite kind of environment uh in our regulatory architecture uh but it's one that's also pragmatic uh so it but you know in it comes down to being able to identify those needs and of course you know Australia is also you got one of the highest population growth rates of any advanced economy because we are an attractive place to live and we can attract you know even if we don't organically grow the skills we can certainly attract them and then have that cross-pollination between people who have recently become Australians residents and those who are more long-standing and you know develop that really great base of of capability and I guess that's where the education piece comes in. If we've got these people coming into Australia, perhaps we can sort of feed them into the education system in this space. But what about competition from overseas, particularly when it comes to salaries? We know that you know Silicon Valley is throwing seven figure salaries at the top talent. uh Asgill, I know that you are looking at how to I suppose uh prevent a a brain drain and we have an example, a video to play of one of the students here at UTS who has actually chosen to stay here in Australia. I see my career kind of changing all the time because technology and the whole software industry is always changing. App development is not just you sitting down on a computer and just coding. Um it could actually fulfill a lot of different parts of me. So the design and the research and connecting to people, the storytelling, that sort of aspect of it because um app development is a lot more about kind of selling your story and selling the hard work that you've kind of come up with. I think I just want to be in a space kind of leadership and also just mentorship for the younger generation of um software developers, designers, UX designers, um researchers. Um I want to be that person that brings a lot of experience um and teaches that to the next generation. In the Apple Foundation program, we are bringing a lot of um women into app development. So we do a lot of programs around um girls high schools um to help them learn about app development and what technology can be and what a career in technology can look like. So do I see myself going overseas to further my career? I would say well the straight answer is no only because I have established my roots here. I've been you know working with the community here. Um, so I want to, I guess, use my resources and use my skills to, um, improve the industry here. I can't help but look overseas, um, and see there's so many other opportunities out there. Um, but I think the main, I guess, foundation of my career is helping the communities here. A lot of pride in that young girl's voice. She wants to stay here. Is that part of it? Absolutely, Alicia. I think uh it's important to um I think people want to contribute to the community and apart from just a financial gain there is a sense of purpose as well which probably we can hear through her uh words. So I think purpose-driven u is a kind of a way the people are uh maybe and that drives a career but I would also like to add that u uh going proximity might not be an issue in future I'm looking at 2030 bjone uh people can work from anywhere and people could probably exchange knowledge and I think it's not necessarily a bad idea to going overseas for resetting I call that resetting you go reset yourself, learn something, come back. But if we talk particularly about US, I mean I know some of my friends and colleagues they moved over there recently. I think uh as overall I have been all over the world. Australia is the best place to live. I have to admit that I've been pretty much in last 20 years everywhere part of certain countries but uh uh but mainly and I I believe that it's it's okay to go learn and come back and then probably introduce bring that knowledge here and build the bridges uh and you can look at that also in a way there might be a really attractive over in salaries or or or payouts you know sense but look at the cost of living over there and what you have is trade-off, right? It's a business. So, I I would say that that's a conscious choice. And when it comes to that knowledge uh that people bring back that it's at the uni, how do we commercialize it? Dr. Terry Roach, you have experience of doing just that. One of the great things about software is that it's being a digital product, you can sell it anywhere in the world. And and in fact, in our experience, we were actually more successful early on outside of Australia than we were inside of Australia. So I think that's a a really a good characteristic and an opportunity that we have in this space. Um, of course I think it's not easy. Um, I think when people think about becoming a software founder and building a business, they typically think all about the product and can they build something compelling and great and exciting and so on and and in fact that's the easy part. Uh the real hard part is to take it to market and find out how to how to actually scale your business and the examples of um businesses that have done that successfully from South Africa are from Australia are um quite unique. Um the there many of us that that really struggle that competition is great that you know one of the things about software as well being being something that doesn't have a high barrier to entry. You know, you can set up a a software business overnight and build your build your technology on cloud systems that you just rent for for peanuts and you don't need large capital investments. But it also is highly competitive just because of that. And there's people all over the world doing that. And so you you really have to try to rise above the noise which is a challenge. Um but so the these are some of the the hard slogs that you need to go through. It's a very great community. we have um you know all of the the um the the founders and the technology uh uh vendors in Australia are are a very close community and support one another and so on. So I think there's definitely great opportunity um but it's it's um it's highly competitive. Yeah. And we talked a lot about the sort of not lax regulatory environment but government wasn't over the top of it when when you know you were starting out and and when these startups were happening but what about incentives David Masters I know that commodity miners for example have incentives should that be replicated in a similar way for this industry yeah I mean one of the points that I've tried to make we should treat IP like we treat minerals in the ground in that um you know we we want. I'm supportive of the mining industry. It's done incredibly well by our economy and and created the standard of living that we have. But but we think about the minerals on the ground. We we want people exploring. We want them to find them. We want to them to exploit those minerals because then we get a benefit. You know, they generate revenue, they generate profits, they generate tax. Um we should be thinking about IP in the same way. So how do we foster the creation of the IP, the registration of the IP and the scaling of the IP more importantly out of Australia? And I think we've got to a point and and you know I think the point that that Lucy made like in a very short space of time we've created a pretty massive export sector that we can now build upon. Um we're a we're a an industry that operates on scale. And so scale begets scale. And so now that you have operators that operate at that scale in Australia, you've got an ecosystem. You can start to see ecosystem effects. So you've got people who've built commercialization skills. So um you know I'll give you an example of a type of of of of skill set that we had to import before that we're now building domestically. So um product managers. So they're people that have expertise in taking a product to market. So working with software engineers to build to meet market fit. And so that was a skill set that didn't really exist in Australia and we had to bring that in from the US. But because we've built that expertise with imported skills, we now have pipelines of um product managers that come in as a grad and enter a product management stream and we've built those people from you know supernuts in Australia. So so I think once you've got that sort of scale operating and then those people that have learned to operate at scale in Atlassian can then go across and help others to scale. So if you go into the software engineering workforce in Canva for example, there's a lot of exit lassians there because they learned their trade craft with us and they've taken that to campa to help them scale. So there's an ecosystem there but it needs to get bigger. Is that how you feel as if I think certainly and I agree that from a university point of view as well uh there has to coexist with our ecosystem because there should be a demand for people to absorb and uh through jobs through setting up new startups but at the same time I see the need in the universities which we probably has started uh J about having a sort of software startup rather than just software development. and project mindset to more product mindset and that brings what Terry mentioned about that the commercialization stuff. So we have sort of started now in our software engineering uh program in a way to so it's not about project delivery it's about also the product end commercialization IP protection the market segmentation so I believe we have to think about the ecosystem when we even teaching as well one thing I want to mention here um uh I I do feel the need of uh stress on the digital infrastructure I think we need to have a strong digital infrastructure uh it's great to have a physical uh infrastructure the way we have but I believe that's we need that would enable software development all the way from somewhere in the remote area to CBD uh I think people don't need to come to the city people might be the new hubs would be suburbs in my view so the ecosystem would be federated sitting in the suburbs and those would be that tiny uh I would say that the powerhouse for developing software rather than the CBD And when we're talking about that infrastructure, Lucy Ellis, I note that, you know, there's a lot of investment often in hard assets. We're going to still need those in a digital world. What's it going to look like? What financial levers are going to need to be pulled? How are we going to build that infrastructure? Well, before I get to get that, I think we need to make a really important distinction here. um sort of traditional software, you know, your apps that, you know, the kind of products that um you've been making, they are a capital light uh business in the sense, you know, you can rent your cloud storage, you know, you need a few laptops, but it's a pretty capital light industry and so the barriers to entry are low. That means you do get entry, you do get competition, you do get innovation, and you do get productivity growth. where you know there is a little bit of a concern to be raised is of course we're now moving into an AI world where it's actually extremely compute inensive you know the the data center construction phase like part of that is you know for cloud computing which is actually capital saving you don't have to have on premises servers uh it's actually energy saving from the perspective of a country like Australia if everyone went to the cloud instead of being on premises But increasingly, you know, you've got AI, you've got, you know, an oligopoly of, you know, the foundational models in the same way that, you know, in the social network era, you have network. So, anything that has network effects or very large sort of initial capital investments inherently has barriers to entry and therefore you don't have the entry, the competition and therefore you don't get the innovation at the same pace. overall as a rule of thumb that you would where you've got, you know, in some ways, you know, a harder industry. It's hard to compete when there's so many other people um so many other people entering where you've got to really compete for space and but that means you've got to keep competing and and build that better mousetrap that better um that better app. So in many ways you a switch to a more oligopolistic model that's more capital heavy firstly that does actually mean we we are you know I mean I'm not concerned about the funding for that physical investment because the amount of money that is being thrown at uh you know by the hyperscalers at data center production whether that's in the US or elsewhere I mean some of these facilities are just enormous but there is actually an implication for the level of competition and the level the growth rate of productivity by pivoting solely to something that's dependent on either network effects which is the the drag we've had for the last decade or so and andor something that involves you know and a massive balance sheet because you've got a massive physical capital uh requirement as in the hyperscalers. So, you know, let's hope the traditional software industry continues to thrive on that basis because that's actually got the fruits of stronger productivity growth. Do you see a polarization then between the really big players and we can talk about the mag seven on the, you know, on Wall Street and people trying to crack this nut. Now I think where we're heading is the sweet spot will will be using those AI tools as a service or as an enhancement of the original software idea. And we're already seeing that I mean obviously you know you Can at Lassian but you know all of those sort of you know traditional software software as a service in the cloud offerings is adding AI capability and they're doing that because that's where the foundational model is now and again the you know history economic history will tell you that in the end the the end benefits are actually reaped by the users and you that might be the software you know the traditional software providers using the AI tools or it might be the end users. But you know we saw during the you know the first computer wave and the internet wave in the late '9s you know we saw a huge amount of investment go into spectrum and go into you know cabling you know fiber optic cabling. Not all of that got used initially. A lot of it was sort of overinvestment and eventually it got used and you know already in financial markets people are talking about you know are we over are we overdoing this you know already is this you know is this 1999 already again I mean and and I mean bubble's kind of an unhelpful phrase it's just are we in an overinvestment phase that we're going to have regret because you know are is the value going to be reaped by that oligopolistic hyperscala sector or is it going to be reaped by the end users history would say the end users and and you know for our living standards. We should hope that's the case, but you know, history may not always repeat. David, you've been nodding profusely during Lucy's um piece there. Do you have experience you can add from Atlassian? Yeah, I think you know where we sit in the sort of application stack. I mean, we are sort of the benefits of that capital investment, but we're not taking the risk ourselves in that capital investment. So, there's a lot of money flowing into that. There's a lot of innovation occurring. We can take that innovation and figure out how useful it is to our end users. So I think that that's the sweet spot for us and I think actually it's a sweet spot for Australia. There's there's there's we're pretty good at figuring out how to use things. I think you know we were talking just before Lucy that we're kind of a problem solver sort of society and I think that's where um as this capital is being invested a lot of it's you know going elsewhere. It's not it's not just in Australia. Um there's a lot of you probably the the numbers in the US are astronomical but we we'll get a proportion of that uh investment. And to Lucy's point, like I think actually we should be attracting as much um data center investment as we can get and and that we can efficiently deploy because we'll utilize that infrastructure at some point. Um and you saw that um that sort of buildout that happened in the late '9s, early 2000s. I mean that capacity was utilized. Um and in fact actually often um some inefficient business models that that you know fell by the wayside led to actually new business models that emerged during the sort of 2010s for example. Um so I think that there's an important piece where we should be attracting as much of that capital as we can get um because that infrastructure will be utilized but I think where the real sweet spot is is figuring out how to utilize those underlying foundational models in user scenarios and that's where we're focused. M well so that's sort of the physical aspect of it and the capital aspect but when it comes to AIS as if how do you see it being used as a tool for your students and then in the real world when we're talking about software developers uh certainly um so first of all we do need technical foundation without that I mean AI could be used for generating code probably specs uh for certain other purposes. Uh automation, wipe coding, people look at that all good, but universities need to put a probably emphasis on that the technical foundation. What if something goes wrong? And actually that's where the software engineering and technical skills comes into play. Can you troubleshoot it? I mean someone can I call citizen developers can develop code or generate code using automation tools. But when it comes to troubleshooting, you need people who can go beyond that large language model. So we do need basic skills. So I see that technical foundation. We're going to be keep doing that. Uh but also we're going to give students before they get to the market to be able to use those commercial tools AI tool. So they keep they're ready. They're very job ready. They could be uh uh they could should be able to use those tools and be be productive. So those things when they do on boarding they don't see that as the foreign thing. So we have to balance at the university uh technical foundation with the AI engineering but AI engineering I would certainly say that uh uh is going to taking a leap in next three four years uh and then part of that would be also uh we have still cobalt programs right mainframe still there. So I was talking to another bank and uh they are going through a legacy code migration and that's where AI could be used to generate specs out of that uh so we can understand uh because for me software is not a piece of code it's the actual business if you look at the financial services the banks are simply a software if you look at the public sector another bigger consumer of the software they don't sell card of commodities the way we go to the retail they are a software so I I believe that in that sense we needed to run two worlds it's pretty much like when you are onboarding a new software you have to run in parallel the traditional legacy software piece as well and we have to do that way uh software engineering point of view so uh so we we want to make sure we have people coming out with solid skills people should be able to evaluate software generating ing code is maybe a 5% of the whole thing but we need a 95% of other stuff which is writing specs. I mean you can use genai but someone will be accountable for those specs and then code can be written using geni against those specs but someone need to validate whether code meets those specs. Um I I was just looking at one of the videos. I'm going to wrap up on that. Um uh uh prompt engineering is dead. Specs is everything. I I really like that line uh a few days ago in AI engineering conference. Terry, what's your experience? How did you scale up the business so quickly? It wasn't the world of the AI sort of, you know, when you started, but it's here now. How would it have changed things for you perhaps? In our case, um we as a small business, we we took a the path that is at least traveled and we tried not to take investment right up right up front. Um we got a little bit of grant money from the government which was really useful and valuable but we realized that the best money that we and funding we should get for our business was um customers. So we penet we the product that we produced had a global community of practitioners and we got deeply involved with those practitioners and got to know them all around the world and through them got into very large global logos and brands adopting our software. So it was it was almost by infiltration that we did that. We got incubated by a few of those large organizations that bought into our idea and and we managed to to grow the business in that way. Um I think the if I think of what's changed and and and what I see right now with the promise of AI um you know change change is opportunity but it's also disruptive and and this is exponential opportunity and exponential disruption right now. It's really difficult for a software vendor right now. We're all scrambling to try and catch up and figure out what does it actually mean? What do we do with it? How do we put it into our products? Um, and just as soon as we get a plan, the next generation comes out and we've been punched in the face as as Mike Tyson says. Um, so we I I think that's the situation. I think what we're seeing right now is a is a period of really um turbulence while we all work out how we transition through this and what the the next generation of software is going to be. I think um there's huge change coming. I think the question of centralization around those consolidation around those those few um organizations that really own the foundation of it all is is is a huge risk because anything you do on the periphery could quite easily suddenly be swallowed up by by them. So you really have to go back to your expertise like software is just a mechanism. What what is it? What's the problem you're trying to solve? What's your domain? Be a domain expert and that's your moat. You know that the the the the software vendor is never going to become domain expert in your domain. Just we what we need to do now is not figure out how to do what we used to do better but do things differently with this new software in to address the problems of our domain. And I think if we approach it that way, we're going to find great opportunities. M and I've got some great questions coming through here, but Lucy, before we get to them, I just want to ask you, how do you see AI improving productivity inside an organization in this sector? Um, great question, Alicia. And there's sort of the micro answer to that and well as the macro one and and just by way of example, the micro uh answer is I actually report to the head of data, digital and AI for the institutional bank at Westpak. So the economics team is considered part of that you know data insights AI and we are all up for you know AI and you know my my my boss actually tried to do a bot well quite successfully actually not just tried to he made a little bot trained on our economics research which of course he called let's ask Lucy uh and um you know so we're we're all for that and you know already seeing you know great kind of you know alongside you co-pilot that whole thing of finding the information, you know, how do I, you know, you know, what, you know, how do I find the terms and conditions for this product? How do I help this customer with this need? You know, what does the economics team say about this industry that I can then feed back to my customer? There's there's huge sort of that enterprise search collation of information. It's enormously powerful. But Terry um pointed out something really important which is again and again it's a lesson of the '9s as well. um it took a while to really reap those productivity benefits. You know, there was actually equipment in 1987, Bob Solo, who was a a very well-known um academic economist uh commented an article that you could see the computer revolution everywhere but the productivity statistics. It's now known as the solo paradox. And that was true in 1987, but by 1997, you could definitely see computers in the productivity statistics. And you know again there was a massive productivity boom productivity growth boom right around the western world and indeed Australia's productivity growth at the time actually outstripped the US's and a lot of people don't realize that this is this is the lesson that using the technology gives you more benefits than being the technology. Uh but you don't see those benefits until you change your business model until you change your business processes. It won't be enough just to say you know help me find that thing. it's going to be you know how do we completely remake how we do things and I mean this is something you know you know my team and I ourselves are grappling with right now and I suppose that comes to the agility point that we need to teach in unis uh David masters you were saying around the critical thinking piece and I'll get your take as well uh as if but I'll start with you what do we need to teach students yeah I think it actually puts a lot of importance on the fundamentals because um if we're going to be relying on um AI agents and LLMs uh to be doing certain functions, then we need to interrogate what it is they're doing and we need to understand what it is they're doing. So, you're going to need people with those critical thinking skills. I actually think it it kind of goes back to the core purpose of universities, which is to teach knowledge and to teach critical thinking and to teach engagement with with information. So, you know, I think where AI is going to have um a lot of impact is actually in process automation and taking things that that are really complicated to do and and simplifying them. But you're still going to need people to sort of ensure that the output of those processes is accurate. Um that it's relying on on data that's that's um that's uh you know been interrogated and and and doesn't have errors in it. And and just even just kind of like you know One of the challenges that I think you'll have with AI is humans have this sort of gut instinct for things that's really hard to train models to do. And so sometimes you look at something and you're like it looks right but there's something about it I'm not quite sure about. And so that causes you to ask a few questions and interrogate things. And so it's hard to train build that into a system. Um but that's where humans will be really important. And I think if we think about like just I mean you're seeing this in social media as well. You know, the critical thing is to teach children, young people, students to really think about the sources of information they're getting to look at, you know, is it is it related to a um, you know, can they find the source of that material? Can they interrogate that? And I think AI is is is similar to that sort of process. We're going to need AI literacy in the same way that we need media and social media literacy. So, ASF, what is the university's approach to getting graduates business ready in this new world? And I know we're only at the very beginning of it. I I think this is a very good question. Uh how you make them ready? I think the foundation year which is probably teaching you u the foundation technical skills and then building onto that you have probably developed expertise in one of the areas because soft engineering is a big thing in a lot of places and then the approaches we have at the moment we have industry studios or our internships or capstone all these things happening currently uh the way we are making them ready because they get a taste of that what's feel like to be a part of a pressure project planning deadlines managing stakeholder communication solving a problem because it's not about writing a code it's about solving a problem through the code uh but I also see beyond that would be there would be a uh school to the profession they're going to be a I think there going to be a three four years of direct you're going to have a study at university and then you're going to be waiting for the job I don't see that going to be uh in the future So that's what we would happen. People might be from school, they go to uh the workplace and then based on their demand they will be probably working with the university. So I think they're going to be parallel kind of a learning and uh doing things going to happen uh uh the way uh the things are going in future u and at the same time I would say that uh uh it's very important to upskill the staff as well internally. So they should be able to teach what is required in the industry. So I think that's where university need to have a partnerships uh to upskilling the staff as well not only students. Uh and then the last thing is really important. I'm going to go back to the infrastructure. I think we have built physical labs in the universities. You can look at that. I think we do need a digital environment, the cloud, the connectivity, the compute, the GPUs uh to be able to have those labs seted up there and the university premises can be or facilities could be used for doing project and working with industry and your teacher. So the fundamental education might be shifting into the cloud as a delivery point of view but actually working on project and stuff that could be on campus. So I think there's a whole operating model going to shift uh as you probably could u imagine uh the way we're moving forward in next three four years. Thank you. Well, we've got the audience questions here and I'm going to throw this one to David simply because you work for a uh a global company that also is listed in the US. Do you think one reason why we see more AI progress in San Francisco than Australia is because of the AI regulations or is it because of talent or are there other reasons in your view and um panel members feel free to pick up on this too? Yeah. Yeah. I think it's it's the depth of the talent pool I think primarily. I mean I think if you look at um I mean that's one of the points we're making you know the the the the importance of having an ecosystem here and and if you want to retain talent in Australia and not have it go overseas then you need to ensure that people have options and so that's not one company that's multiple companies and so you know one of the challenges we had in bringing people to Australia initially was they might come and work for us but where else do they go like so if it doesn't work out with us it's a big risk to move you and your family across um you know from the states or from Europe to Australia, you want to ensure that you've got options and so that's why the valley has always been so successful and you've seen that with AI talent now because you've got this incredible talent pool. Um you've got incredible expertise that's been coming out of institutions like Stanford over a number of years. Um and so I think that just um density of talent, density of capital as well. um destroying I I think it's less to do obviously you know sort of um the US has sort of taken the foot off the regulatory um um break completely um which is also stimulating things but but I don't think that's the core factor Terry or Lucy yeah well I just comment that this is just absolute standard economics that Paul Krugman won his Nobel Prize for you know not quite 20 years ago which is elomeration people go where the jobs are and then the firm firms go where the people are and it feeds on itself and so that's why you know London is still important for both media and finance New York is still important for finance you San Francisco very much you more for tech you know you want to be in the country music industry you move to Nashville you know the elomeration economies feed on themselves and so it's always going to be hard to break out from that but ultimately you know the cost of living in a big expensive city also builds and so there is a trade-off there. And so where I think you know Australia and and I wouldn't just say Sydney. I think you know we we forget that there's some some interesting stuff going on in Melbourne uh and you know we actually have a a big unit of um you know developers over up on the Gold Coast as well. Um but you know where there's also ancillary you know the fact there's a big funds management industry in Melbourne there are sort of ancillary industries where developer talent is also useful you know you know financial services both in Sydney and Melbourne uh is also part of the story and let's not forget defense tech either because that's a really software intensive integration intensive industry uh and you know we also know that that is also diversifying out of the US now because of the geopolitical environment and we've seen quite success with drone shield. I mean that stock is I know that's not how you measure things but from a financial markets watcher that stock is up uh 570% this year so far. I mean they've been very successful very quickly. We'll never ask financial advice from a macroeconomist. But you know this is this is an example of the point where if you have the right sort of educational levels and pragmatism around problem solving little niches will come up and it's not really you know do we have this industry or that industry the way a regulator is inclined to think about it is can we pivot and I mean we've just seen this with Ukraine. I mean, who knew that Ukraine would be an incredible drone industry? Well, that had to be. Uh, but you know, the ability, it's the ability to pivot from where you are to where you need to be. That's the fundamental skill. Yeah. Terry, do you have something to add there? Yeah, I was going to say, I mean, it's so clearly the mecca, the San Francisco, the Silicon Valley for for software. Um but but and and and all the points made I won't try and repeat but I think the other major value in in the temptation for somebody like myself who once you've started here and built some software here is that the the the challenge is scaling and and those are the people that know how to do it. you know, all the talent, the people that have done that, the funders, the the the you know, when you go and get if you're looking for funding, yes, you can find funding over here, but you can't find the expertise from those investors that you can find from the investors over there. So, it is a real challenge and and a huge temptation uh for for um tech. Now what about the importance of AI and data governance and regulatory and safety? I I think this question goes to uh partnerships with the Asian region. It says the Asian global south century. Um what about government framework and cooperation in that space perhaps that is um for you David? Yeah, look, I I think um you know, we talked about patchwork regulation before and I think that creates challenges and one of the one of the the things that that I think is incumbent on um our government to talk to other governments that the more you can collaborate, it creates opportunities for companies that emerge in your jurisdiction to scale more efficiently. So the more regulatory barriers that each country implements, the less opportunity it creates to scale. And I think if you look at sort of what's happened in Europe over the last sort of you know 10 20 years there's a real conversation around competitiveness of of European companies globally and and part of that is Europe is big enough to be able to build a business in Europe and not think too you know too broadly but also because um there's so much European specific regulation you design your company to meet those regulatory requirements that may not mean that may mean you're less efficient when you operate outside of Europe. And so if you look at software as a really good example, Europe hasn't created two businessto business software companies of the scale of Atlassian or Camber in the last 20 25 years. You know, it's a market of 700 million people, which it seems really odd, but but I think it's a combination of some of those factors. Did you have something to add there? Yeah, I think um I think Australian Australian government gets um a bad rap uh for for their lack of investment in the sector. I would be adding to that perhaps in some way, but there there is a lot that they do. Um and I think um so we we benefited from the um R&D tax refund which was very very good and it helps to compensate for the cost of of of um people down in this part of the world. But I think one thing that is really disappointing that that they could do much more of is make it accept. Yeah. Government is a huge customer for software and particularly the type of software we had and it is just so hard to sell to government for a small business. you you can just break your your business um trying to go through the the the the the navigate the processes of of selling to government. You can win tenders uh and then two years later they still haven't spent the money with you, you know. Um so it's a it's a it if they would just buy our software uh that would be a great help. I know David that um you have pushed for the R&D in uh tax incentive caps to be lifted from $150 million. The BCA wants something similar. How would that help? Uh well, look, it's I mean, if you look at capital um will flow where it's most efficient to deploy. And so if you're uh if you're thinking about where you're making your R&D investments, you think about the regimes that operate and and and there's a combination of factors like how expensive is it to hire talent? can you get access to the talent and then once you have the talent you know is it efficient to continue to employ them in those jurisdictions and so I think the one of the challenges you with the R&D tax incentive being capped is it it kind of means that that once we go past that cap you know R&D engineers that we hire in Australia become less efficient and so our point has been look if you want companies like ours to continue to scale in Australia then you need to create incentives that allow us to continue to scale and so an artificial cap is is problematic in that regard. You know, I think, you know, the R&D tax incentive has been super important for us to be able to build our R&D engineering workforce in Australia. And when we talk about the sort of the the scale of talent, um that um the report that we worked on with the BCA sort of talked about the importance of having those large players in the ecosystem. So you can trace from the top 5% of R&D tax incentive claimants about 1,800 CEOs and founders um across the economy. You can also then uh look at at how that um those skills have been deployed elsewhere. So you can trace from about top 5% um to about 8,200 other um uh you know um companies across the economy. So they're an important feeding ground for talent. And so if you don't have that scale then you don't get those ecosystem benefits. And that's the point that we're trying to make to to remove some of those artificial barriers to continuing to scale those operations. It's us, it's Coccia, it's CSL, it's ResMed, it's those companies that you want to continue to scale here. Okay, we're getting to the end of the evening, but um as if Gill uh I have a question here. Uh what can students do if we see better opportunities in the US or Europe? People would be there as uh Australian players. Um, but it's slow in adapting to AI especially. I'm I'm not exactly sure what this questioner is getting at, but I wonder if they're just sort of positioning where the opportunities are. Should they stay here or go? I think they should go where the opportunities are, but I think the question I'm hearing that read between the lines. Certainly there is a very slow adoption towards AI here. So if you people want to do something uh interesting they actually look for the place where they're being adopted and I think US is one place uh we have Middle East as well um and in Middle East different countries come together and then they invest a true trillion dollars into that to creating AI hub. So uh I think if we create opportunities here as well and partner if you can't I mean we are comparative small country if we partner with others rather than try to do everything ourselves either US or Middle East somewhere or or neighbors are in Apac I think that's what we need to look at that how we can have uh the talent retain here uh but also it's okay to go over there learn and come back so you can you can bring that mindset I think uh the mindset that is in a way is really really important uh uh because either we are complacent to what we got it I think we got everything in a way uh or we going to challenge ourselves so I think I think that's where uh things come up and I think I'm going to pick on the funding side as well uh in increasingly we the government is reducing investment in high education in R&D I think some of the work we did with our industry partners and they're quite big now we started small they're big but then went to US later on uh because we started with maybe seven eight people now they went overseas so there's a there are many case studies by the way through through uh through UTS work so I think the funding the sustainability more how you scale and scaling is an option I would say that funding point of view or incentivize large corporates I going to say that uh the kind of people they hire they are more from not the deep tech they're more from the solution side and there are not many R&D labs here so I think this is important those large corporate needs to be partner incentivized beyond tax to be able to bring innovation here so that people can expose to that innovation uh proximity being not a challenge because we can connect them remotely so a few things to change well one final question and I'm going augment something that one of the audience members has submitted very quickly because we've got less than five minutes so about 30 second answer each. What change does the panel think is needed to scale the ecosystem here to achieve roughly $20 billion in software exports? Uh I might start with uh you Lucy. Just keep mining the good ideas. Terry, I think if it's ever going to happen, it's going to happen now because there's so much opportunity right now. So, I think um get out there and innovate and and and we can do it. Uh just believe and and it's not difficult. It's not going to cost you a lot. There is a a fertile market out there for making your ideas something tangible and digital as if I think ideas for sure as starting point, but I think enablers needs to be there. enable I see that the strong ecosystem and digital infrastructure if people do not have GPU to process the models there's no innovation right and if you ask the universities around there how much GPU power they have or do they have their own AI cloud I'll be surprised to see if someone says yes David do you have one change that could be made uh or a couple yeah probably a couple um look I think we just need to to look outside the bubble of Australia and think about like are we competitive with other markets in terms of attracting um well incentivizing the investment in intellectual property the registration of intellectual property and the scaling of intellectual property so if we think about there's a range of settings involved in that and talent is part of that as well so um it's not one simple change but I think it's just basically looking at where we sit from a competitive environment and are we competitive in attracting that capital and that investment so so much deep knowledge and insights it's an absolut absolute privilege to have had this conversation and I very much look forward to as a financial journalist uh leading the news bulletin with not BHP and Rio Tinto but our software developers as the next frontier for software experts so uh exports rather but to the experts David Masters thank you so much for your time tonight as Gil we appreciate your time as well Lucy Ellis thank you so much and Dr. Terry Roach, thank you so much for your time. If you join me in giving them a warm round of applause. [Applause] So I believe that post this event there will be an email to go out which you will receive in your inbox where you can provide feedback for this evening's event and offer various comments. Your feedback is really welcome and I'm sorry to the audience members whose questions I didn't get to tonight, but I think it's safe to say we have certainly learned a lot from our esteemed panelists on the stage here with us tonight. [Applause]