Parliamentary Committees

Select Committee on Artificial Intelligence

Mr BROWN (Florey) (11:02): I move:

That the report of the committee for the Fifty-Fifth Parliament be noted.

On 6 July this year, I moved to establish a select committee to inquire into and report on the current state of artificial intelligence development, deployment and application across various sectors, with a particular focus on the economic, social and ethical implications for South Australia. In my speech to the house on this day, I acknowledge that we are on the precipice of major change and that artificial intelligence—or AI as I will refer to it henceforth—is fast becoming part of our day-to-day lives, in our homes and in the workplace. Like it or not, the age of AI is upon us.

I have previously stated that there is an opportunity for this house to ensure the proper framework exists so that our state can continue to be a leader in AI research and application to enhance all our lives. AI is a rapidly expanding field and one that we have very quickly come to rely on. AI technology has integrated itself almost invisibly into the everyday lives of South Australians, from the use of virtual personal assistants such as Apple's Siri or Google Home, to generative AI such as ChatGPT, to everyday apps such as Uber and Google Maps, or even tech devices such as robot vacuum cleaners.

Not only after hearing the broad range of evidence presented but especially after hearing it, I believe that we are truly at the beginning of a new period of history. The constant announcements of new advancements must have been what it was like for the Victorians during the advent of the age of railways. Both technologies prompted rapid, disruptive change that transformed economies and societies. Just as the railroad provided faster and cheaper transportation of goods and people, AI promises to enable faster and more efficient completion of all manner of tasks.

Railroads fuelled the industrial revolution by expanding markets and integrating supply chains. Similarly, AI stands to boost productivity across all sectors of our economy, from manufacturing to services. For example, AI can optimise complex mining operations and port logistics, augment health care through improved diagnostics, and assist financial firms in detecting fraud. Adoption of AI in agriculture can increase yields and reduce costs, boosting the competitiveness of South Australia's food and wine exports.

Both railways and AI generated excitement about their transformative potential but also fears about their disruptive impact. When railways expanded across South Australia in the late 1800s, some feared it would disrupt traditional pastoral ways of life. Many early Victorians believed that travelling by train led to madness. Today, some view AI as a threat to human employment and agency. Others believe its use will mean the extinction of all human life.

At the risk of stretching the comparison, it should be noted that there are key differences between the two technologies as well. The railway age triggered centralisation of industry and population in Adelaide. AI enables decentralisation and remote work, potentially revitalising rural communities struggling with youth flight. While railways supported the development of large corporations, AI has the capacity to empower smaller firms and independent workers to better compete. The economic gains from railways took decades to be fully realised, but AI adoption is progressing rapidly, promising quicker enhancements to efficiency.

Intelligent embracers of AI today can boost South Australia's competitiveness. Our state was at the forefront of railway expansion in Australia. Early adoption of AI could again give South Australia an edge in attracting high-tech future-oriented industries. Yet, we must also ensure the economic gains from AI are distributed inclusively. As with industrialisation, government has a role in assisting displaced workers through retraining programs. State initiatives to provide affordable broadband access can democratise the utilisation of AI tools. AI's transformative capacity across so many sectors of our social and economic lives might be said only to be limited by the human capacity to imagine it and the intentions and ambitions of those who seek to develop and implement it.

Our state has already become a leader in the field of AI, boasting success stories like AI-powered talent marketplace Pickstar, event and stadium point-of-sale application developers MyVenue, visual effects studio Rising Sun Pictures, geospatial technology company Aerometrex, media insights company Amplified Intelligence, and open-source intelligence solutions developer Fivecast.

Adelaide also hosts research institutes like the Australian Institute for Machine Learning, which in 2021 ranked second globally for computer vision research published over the period 2016 to 2021, as well as MIT's Adelaide Living Lab, the collective of innovators at Lot Fourteen, and offices of international companies like Amazon Web Services and Accenture.

However, despite its place in our lives and the economic potential it offers our state, many Australians still view AI with a great deal of trepidation, associating the technology with job losses, deepfakes or decreased privacy. The select committee looked at how these concerns can be addressed and how AI can be harnessed for the economic and social wellbeing of South Australians, by examining how it can be appropriately used by government in a way that is ethically sound but also innovative and in line with our state's security interests.

The committee invited submissions addressing terms of reference that sought to examine how SA is using AI now across the economy, from agriculture to mining, how it can be used in the future and what challenges may arise both from its use in these and other sectors and from rogue use.

We received 46 submissions from an array of stakeholders. The committee then heard from a wide range of witnesses, with 21 people appearing at public hearings, representing state government departments, specialist artificial intelligence institutions, researchers and AI experts. Consistent across the submissions received and hearings held by the committee was excitement about the possibilities offered by AI. AI can and does benefit South Australians, whether it be in our state school classrooms helping students learn or assisting SA Police solve crimes to keep the community safe.

The committee has developed 15 recommendations that seek to further this enthusiasm and also mitigate any risks involved with AI advancement, particularly those pertaining to privacy and image-based abuse, as the excitement surrounding AI was also tempered by concerns that regulations surrounding privacy may not be stringent enough to keep up with progress in the field. Therefore, the committee recommends possible reforms of laws pertaining to the use of AI to ensure the privacy and wellbeing of South Australians.

The committee further recommends that these laws be consistent with the rest of the country, keeping SA on an even playing field with the rest of Australia, and that the state government coordinates with the commonwealth to do this. A strong emphasis has also been placed on the need for continual and further investment in the field to foster innovation and economic prosperity.

The notion of sovereign capability was an important issue the committee touched upon, and it believes that opportunities for the successful use of AI across South Australia is somewhat reliant on our ability to generate this capability. This can mitigate security risks associated with using AI technologies developed offshore, lessen the risk of bias for Australian users and maintain self-reliance in an unpredictable world.

The committee is particularly interested in how South Australia can further build on its reputation as a leader in AI technologies and has developed several recommendations that look to continue to foster sovereign AI capability. We are also educating the next generation of workers, exploring how AI can be used more effectively in the classroom and across the curriculum.

The committee considers educating students about AI will not only keep them safe but also foster talent in AI-related areas, thereby building an AI-ready workforce for the future, leading to greater sovereign capability. The committee believes that adequate instruction about AI in how to use it safely and ethically is essential across the educational spectrum and for all age groups across the community, including groups such as older South Australians who may not feel as technologically literate as their younger peers.

Some members of the workforce may also express fear that AI is coming for their jobs, that AI might render them obsolete. However, from what the committee has heard, human oversight is essential to the successful and, moreover, ethical use of AI. The committee found that this human-in-the-loop element is crucial across all sectors, from health care to policing, and heard repeated evidence stating that a human element must remain to ensure human accountability in its use and to maintain accuracy in decision-making.

Retaining this human oversight can also combat any bias in the use of AI. It is equally pertinent to point out that, like the industrial revolution, AI will likely result in new roles being created, including AI-related roles not yet imagined. Additionally, as AI replaces some tasks, workers may be able to shift their focus to other tasks that are less repetitive, or they may have the opportunity in time to undergo further training and build upon their skill set.

Throughout the inquiry, the committee was impressed with how state government departments and agencies are already integrating AI into their work practices. We wish for this to continue and recommend that frameworks are put in place to enable continual success in these areas. I would also like to particularly mention the Department for Education, who chose not to issue any blanket bans on generative AI tools in schools but, instead, have taken a cautious but ambitious approach to make sure that South Australia can get the most value out of the opportunities created by AI in education.

I am reminded of the sentiments expressed by the department's chief executive, Professor Martin Westwell, who told the committee:

…our students are going to need to work in a world where AI exists, and we need to teach them to use that technology ethically and productively.

One area of education where AI shows a particular promise is its use to tailor learning for individual student need, both from the assessment angle and the instruction angle. In a TED talk in April this year entitled 'How AI could save (not destroy) education', Sal Kahn—the founder of Khan Academy and the Khan Lab School in Mountain View, California—spoke about how AI could be used to create a personalised instruction environment, how the very latest in technology could be used to reinvigorate the use of the Socratic method.

He envisages providing every student with a bespoke personal AI tutor whose output is highly responsive to and precisely tailored to their learning needs, informed by the evaluation not only of whether the students are answering questions correctly or incorrectly but, in instances where they are not correct, using models to ascertain how they arrived at their incorrect conclusions. AI will be able to use that information to help them build a stronger understanding in exactly the way they need in the moment.

This is exactly how individualised instruction should work, but because it is prohibitively costly for every student in our schools to have individually tailored instruction from humans in the flesh, the potential for AI to create a state were individually tailored instruction can be provided to every student feels almost within our grasp.

The committee is also aware of the importance of effective governance and human oversight and has developed recommendations that seek to build on existing frameworks and also collaborate with state and commonwealth governments to ensure consistency of legislation as it pertains to AI. This has led to a recommendation that the government consider the formation of a permanent, whole-of-government body or group dedicated to creating and implementing an AI framework based on accountability, transparency, societal and environmental wellbeing, universal access, fairness, safety and security, and human agency and oversight.

While South Australia has quickly established itself as a leading destination in the field of artificial intelligence, the speed at which AI technologies are developed is currently working at a faster pace. The committee is mindful that, for the state to maintain its reputation as an AI hub, continual significant work must be done to keep and attract research and investment. Furthermore, rather than be the proverbial jack-of-all-trades but master of none, the committee recommends that South Australia build a competitive advantage by identifying selected areas of AI to champion and develop world-leading levels of expertise.

To achieve this, the development of a state action plan is recommended to effectively market South Australia as a premier destination for AI investment in Australia. Support from the Department for Industry, Innovation and Science to harness critical technologies such as AI can also provide a pathway to drive both innovation and South Australia's economic growth in some of the state's key strategic economic areas.

One of the challenges of this committee was to keep abreast of developments in the field that pertained to our remit—no humble feat, given the speed at which AI innovation and technology are advancing. Indeed, as the final touches were being made to this report, President Biden issued an executive order establishing new standards for AI safety and security to safeguard the privacy of American citizens and protect their national interests. Now, as I speak to you today, some technologies mentioned in the body of the report may already be obsolete or superseded by a new, updated product, such is the pace at which this field is moving.

It is the hope of the committee that the recommendations contained in this report acknowledge and allow South Australia to keep pace with the rate at which AI technology is advancing. In dealing with such a complex and malleable topic, expert advice is essential. The committee thanks those who submitted to the inquiry or attended a hearing and values their contributions in developing these recommendations.

I would also like to thank the committee members—the member for Davenport, the member for Elizabeth, the member for Heysen, the member for Morphett and the member for Narungga—who have shared my interest in this important topic and have collaborated to form recommendations that we believe represent constructive steps forward on this complex matter. I would also like to thank the committee staff: Melissa Campaniello, Dr Amy Mead and Hadi Slyman.

Lastly, I speak to this report today and note that the National Artificial Intelligence Centre has commenced Australia's AI awareness month from 15 November to 15 December, an exciting initiative that looks to illuminate the brilliance of Australia's AI landscape. I encourage my colleagues to look into this initiative and possibly participate in one of the events on the program. I commend the report to the house.

The SPEAKER: That was most interesting, member for Florey. You had me at railways.

Mr PATTERSON (Morphett) (11:16): I will aim to continue that interest in this important topic. I also want to speak on artificial intelligence. I was a member of the select committee, as was mentioned by the member for Florey, and, as you said, AI has certainly recently captured a lot of interest. It is a rapidly expanding field and we are seeing a lot of applications that are now becoming ubiquitous in our everyday life.

So many people would use Google Maps to get where they need to go, especially new destinations; gone are the old days of having to have the Lonely Planet guide to get around in a foreign city. You can just take your phone and you have Google Maps which really gets you to where you need to go, especially in highly populated areas where the application is very accurate. You also have Uber, of course, and people now find it very easy to get around and can order transport wherever they are.

Then we are seeing it come into our lives with these assistants: either Siri in the case of Apple, or Google Home. Most recently, ChatGPT has really stirred things along, which is generative AI which has really taken the interaction of AI to the next level. While it might seem that this has happened quickly, it has been slowly building over many decades in terms of the work. There has been incremental change and then all of a sudden it seems to have pushed over the dam wall. Really, we have seen it move from science fiction into our everyday lives.

Of course for people of my generation, probably their first experience of AI would have come out of The Terminator movies, seeing a future where we saw a computer actually gain self-awareness and launch nuclear missiles and create a post-apocalyptic world. That brings with it fear, as many people see that as something to be cautious about with computers taking over from humans, whereas in reality what we are actually seeing at the moment is that AI can lead to significant productivity gains in business and also in people's personal lives.

There are certainly some risks that we need to be rightly concerned about. A lot of them are based on how fast-moving the technology is. It is very hard for legislators to keep up. In a submission, one of the points made was that legislation can often take years and by then the technology that people were concerned about has actually been surpassed anyway.

As I said, many people are interested in this and, for me, with my background, there is a personal interest as well. When I studied electrical and electronic engineering, one of the areas of study was in computer science and artificial intelligence. As is now, back then the definition of artificial intelligence was that a human using a platform would not know whether there was human on the other side of the screen or in fact a computer.

Of course, in days gone by computing power was not able to cater for millions and millions of scenarios, so it was quite easy to trick these computers and they became quite limited. But as computer power has increased, these processors on the other side handling these applications are able to process more and more data, so can basically appear as if it is a human, thereby introducing it as artificial intelligence. In fact, in my final year of university, I built from scratch a neural network that was designed around predicting salinity in the River Murray. Fast-forward and this form of machine learning is now becoming commonplace and much more advanced based on the computer power of hardware.

In a way, these things revolve around training neural networks to recognise patterns. Effectively, you train them on a whole list of data to then recognise patterns and it becomes a bit like a black box in a way where you are putting information in and you are getting a result out and are not quite sure definitely how that decision was arrived at. That is a caution in terms of AI in terms of knowing how a decision was arrived at.

That leads us to where we are at now with generative AI. It uses natural language to be able to provide outputs to humans. In these early days, because the datasets they have been trained on may be limited, some of the results coming out are based on what information they have been trained on. Most people say that poor data in leads to poor data out and that can certainly be the case at the moment, so we do need to be careful with that as the committee found.

In terms of where South Australia sits in artificial intelligence, we have the Australian Institute for Machine Learning. That has grown out of the University of Adelaide's computer science department's electrical and electronic engineering department. I know it well. When I was going through, they were already looking at computer vision. Fast-forward now and, as was said by the member for Florey, the Australian Institute for Machine Learning is the number two institute in the world in terms of artificial vision. We also have the MIT Living Lab based here.

These are really key pillars. They are based in Lot Fourteen and are really centres of excellence that are attracting global interest into South Australia and bringing with that significant companies. We have had Accenture set up here in Adelaide, South Australia. We have had AWS, one of the top five companies in the world, and Google set up a presence here, as well as Cognizant and MTX.

Of course, they bring with them skilled workers as well, which really just increases the intelligence and problem-solving ability that South Australia has at its hands because a lot of these really intelligent people are problem solvers as well. We also have homegrown companies. One of those is Life Whisperer. They do AI-enhanced fertility, which is a real-world application helping couples with fertility to make it easier for them to have the joy of babies.

We also have MyVenue, which does point of sale; Rising Sun Pictures, which wins Academy Awards for their visual effects; and Fivecast, which is an AI-enhanced intelligent security platform. One of their directors and owners, Dr Brent Cooper, went to uni with me. He is a very smart, intelligent individual and he has been able to set up a fantastic company here in South Australia that is going to be of benefit to South Australia more broadly.

In terms of the committee, we heard from a number of key stakeholders; in fact, 46 submissions were received. We had 21 people appear in front of the committee, and overall they were very excited about the possibilities. They likened AI to the introduction of the telephone and the internet in terms of the big step change it is able to make from a communications perspective.

Of course, there were some concerns as well around the possibility for this tech to do harm in a number of areas. It can be used by criminals. It can be used in terms of computer generation of quite realistic voice mimicry to trick people. There is also AI-based image fakery. I think just in the news last week Dick Smith was again having to point out that it was not him who was endorsing a quite spurious financial product. AI is being used to mimic trusted sources to dupe people into taking action they otherwise would not. It could potentially be used to mimic politicians as well and cause panic throughout democratic societies, so we have to be on our guard for that.

It was pointed out that some of the ways to do that are through effective governance and human oversight. If I could maybe leave with the comments made by Professor Simon Lucey from AIML, he was very wary, though, of having a kneejerk reaction and overregulating. He says it is important that we allow AI to develop here in South Australia because it is going to be important for our South Australian companies to compete in the global environment. If you have overseas companies who are using AI to leapfrog ahead, so too must South Australian companies.

I thank all other members of the committee for their contributions and say that it was time well spent, and it will benefit this parliament as it goes forward in its decision-making around this important area.

Mr ELLIS (Narungga) (11:26): I would like to add my thoughts on this report and start by thanking this house for allowing me to be on the committee. It really was, as the member for Morphett has said, time well spent and thoroughly enjoyable. A very informative set of witnesses appeared before the committee and offered evidence as to the future and the current state of AI in this state and more broadly. It was a wonderful thing.

For me, the takeaway was that there is no doubt that artificial intelligence is going to dramatically change the way of life for us as time progresses. It certainly seems like we are on the precipice of dramatic change in technology, and the way in which we handle that, empower that and harness that will be of critical importance going forward. That came through in spades, and it was certainly interesting to hear different viewpoints about what we need to do to make sure we are ready for that change as it comes.

There is some fear in the community, I think, about how AI will impact people, particularly with regard to how it might impact their employment and job prospects going forward. There certainly does seem to be some threat, particularly as I see it, in terms of white-collar employment. There seems to be quite obvious overlap in terms of the functions of what we know today as AI and what some white-collar jobs are doing. It will be interesting to see how that progresses and whether those labour-intensive jobs which are difficult to do currently due to cost come back in vogue, and we find the return to manufacturing, inventing, building and those sorts of things as time progresses, where computers are doing the thinking and we are doing the doing. It will be interesting to see what happens as time progresses.

I think we can take solace from the fact, as has been referenced in the member for Florey's speech, that at all times in human history when society has been confronted with dramatic change, humanity has found a way to prosper and thrive. I am sure this will be exactly the same and that we will find an adequate way to harness and build on AI and use it to our advantage. It surely will be exciting times ahead if we do it properly.

There were a few things that I wanted to touch on specifically with regard to the report. Certainly a wide range of issues were discussed, and I do not mean to discount any other things but, for me, there were four that I wanted to bring up in the context of this contribution. Firstly, we heard some really exciting evidence from PIRSA about the opportunities that would present to them, to the ag sector, with the advent of artificial intelligence.

To start with, we heard some wonderful examples from the agtech grants that have been opened recently, the initiatives that were successful in securing funding as a part of that grant program and the technology that is coming as a result of those grant programs and that funding. It will have a dramatic impact on ag. It will help lessen costs and make more targeted action, and ultimately there will be a better bottom line for ag businesses and, hopefully, a better product at market for us consumers. There are some really exciting things happening in that space.

It was identified by PIRSA that the management of snails is something that is very much at the forefront now and not too far away. There was the evidence that ordering of chemicals and inputs would be able to be done more efficiently and with less waste as a result of AI. There is also technology to identify fruit fly, which is the world's largest agricultural trade pest.

So there are some really exciting things happening in technology in ag. I think they will continue to grow and expand as AI continues to grow and expand, and I look forward to seeing how that will impact our local businesses in Narungga and our local ag businesses and hopefully lead to better bottom lines—more profitable businesses—and better things for our ag community.

Another thing locally I really wanted to touch on, which is an area of great passion of mine, is I think in the fullness of time AI will present a wonderful opportunity to help alleviate some of the pressures that are being felt by our regional health system. The technology is there to help diagnose and recommend treatments and those sorts of things. I think it will be a wonderful addition to areas where it is difficult to find doctors where we might be able to empower nurses to make an initial diagnosis and have AI verify or confirm and those sorts of things. It really will, hopefully, be a wonderful improvement to our system.

In a rather significant coincidence, I actually had the great fortune to have in parliament this morning the team from Project Checkmate. At the Paskeville field days recently, Project Checkmate, which is a project initiated by the Rosemary Bryant AO Research Centre and co-funded by Skin Check Champions, set up a stall. They had trained local nurses in the diagnosis of melanomas. So they had a stall there, and they encouraged people who were walking past or visiting to pop in and have their skin checked by a trained local nurse, who was then able to make an initial assessment and then refer that assessment to AI that was on site which would either be able to confirm or deny that diagnosis.

Where in the past we have had pressure put on the GP system and a little bit of a blockage form through the difficulty in getting appointments, we are able to use some latent workforce—the nurses—to make that initial diagnosis and have AI there to support them, to back them up and to confirm that diagnosis. That was a wonderful thing. They had 170-odd people through over the three days of the field days. They trained eight local nurses up to make those diagnoses, and it was really wonderful, a raging success, I think.

I had Professor Marion Eckert, Mr Greg Sharplin and Ms Pam Adelson in today to talk about that and talk about a path forward as to how we can make sure that goes from a pilot program to an actual training program to train our nurses and to improve that AI technology so that it can make that confirmation or that verification and really have a wonderful impact on the prevention of melanoma.

The difference between detecting a stage 1 melanoma and stage 4 melanoma is hundreds of thousands of dollars per case, so if we can make those savings it would be wonderful for the economy. It will certainly be wonderful for the health system, keeping people out of the more critical care. It will be a wonderful benefit all round.

That is just one example. While I am at it, I should urge the federal government and Mark Butler to agree to that funding program. They have put a rather sensible proposal forward to him recently—I think it is on his desk at the moment—where they are working to convert it, as I said, from a pilot to an actual program. It will be money well spent, for mine. Here is hoping that Mark Butler agrees and we can see that project rolled out more fully.

It was wonderful to have them in parliament. I thank them for the work they did at the field days and training the local nurses and treating local people. It is just one example of how AI can help with our regional health care going forward.

There are a couple of other things, and they have been touched on already, so I will not spend a great deal of time on them, but I think it would be really exciting to make sure that South Australia continues to be positioned as a hub for AI technology. We have heard from previous speakers the current situation and current businesses we have been able to attract and the progress we have been able to make but, as the member for Florey said, if we can target our effort and make sure we are not trying to be a jack-of-all-trades and really focus our attention on some areas that are winnable it will be a wonderful thing to have here in South Australia, here in Adelaide—to have some high-tech jobs paving the way into the future. I look forward to seeing how that plays out.

The other thing that is worth touching on slightly more is that we do need to be wary about the pitfalls of AI. There is an opportunity cost for everything, isn't there, and some of the evils that come along with the exciting prospects of AI really have the potential to have a dramatic impact on people. I think particularly of evidence of deepfake images being used for blackmail and extortion, which could be really damaging to sections of society. There are recommendations in this report about training people up and educating people about the evils of AI and how to spot those pitfalls and identify them before they become a problem, which is a really valuable thing and might be of the highest priority in this report going forward. I commend those recommendations specifically, and all 15 recommendations. It really was a tremendous committee to be a part of.

It is a really exciting time in our world's progression, I think, as we watch this technology develop and the opportunities that it will present to businesses and individuals right across the spectrum of industries. I commend the report to the house. I thank the committee for allowing me to be a part of it and I look forward to hearing all further speakers.

Mr TEAGUE (Heysen) (11:35): In joining the reflections on the report of the AI committee, I want to thank the member for Playford—Florey, sorry. I withdraw that. I did not rely on AI, even, for it.

I want to thank the member for Florey for chairing the committee and for his remarks outlining the work of the committee and the report that we are here noting. I also wholeheartedly agree with and will not repeat those observations of the member for Morphett who, perhaps more than any of us on the committee, brings that wealth of technical and professional expertise to the consideration of the opportunities and challenges that AI is presenting at this time. The member for Narungga has reflected on some of those key areas of opportunity in industry that might affect us and help us in each of our local areas.

I resisted, more generally, any urge to seek the assistance of AI in the preparation of these remarks. I think there is a cautionary note for all of us who are engaging in the world of technology, the analysis of data, the presentation of advocacy and so on, that we make sure that we are using technology and deploying it for the powerful purposes that it is able to serve.

It might be well to reflect on the events in New York in June this year. Judge Castel, of the US District Court, was moved to sanction two practitioners and the law firm, I understand, by the name of Levidow, Levidow & Oberman for whom they practised, for what has become a particularly celebrated example of a misuse of AI in the most basic of ways. Those two unfortunate practitioners found themselves providing written submissions to the court including reliance upon six wholly fictitious authorities, as it turned out. Apart from monetary sanction, the judge has moved to require those practitioners to go and apologise to the very real judges who were supposed to have made these fictitious findings and the parties concerned and so on.

It is certainly not as simple as it looks. I think it is interesting to bear in mind what might be the sentiment of many such amateur would-be 'dip the toe in' in terms of use of AI. The law firm, in response, made a statement that included that they asserted:

We made a good faith mistake in failing to believe that a piece of technology could be making up cases out of whole cloth…

That is what the firm's statement said, in what seemed to be an attempt to couch a pretty mundane mistake in rather more elegant language than it deserved.

There is clearly this now wealth, if not universe, of transformative change that will now be before us, confronting us and providing us with a chance to proceed. The member for Florey has couched that in terms no less than the transformative effect of previous industrial and technology revolutions, and I think it is right to couch it in those terms. I am less inclined to perhaps characterise it in terms of a precipice but more perhaps a point of embarkation, a base camp or a point at which we are to see quite comprehensive and rapid change as we go forward from this point.

I make particular mention of the 31^st of 46 or so written submissions received by the committee. The Australian Institute for Machine Learning participated both in that regard and as a witness to the committee; the work it is doing and the innovative activities in this space otherwise conducted at Lot Fourteen are truly remarkable. They are beacons for and ornaments to our state and will show the way in terms of innovation over the years ahead.

Of the 15 recommendations of the report, I highlight that, while there is a mixture of recommendations to industry and for the purpose of public sector to government, recommendation 11 is really directing a recommendation very much on the education side, to recommend that the STEM curriculum in our state is very deliberately focused on enhancing the capacity of those students in this state to get to grips with and take full advantage of the opportunities this new technology will present.

As other speakers have, I commend the report. I hope it is a useful contribution to our tracking forward as a state, and I hope it is a subject the parliament can come back to and revisit before too long.

Mr BROWN (Florey) (11:42): I want to take this opportunity to thank all those who contributed to the debate. In closing, I echo some of the sentiments expressed by the member for Heysen about AIML, who are world leaders in this area of research. I have no doubt that in the years to come, whilst it is certainly the case now that AIML are not known universally in South Australia, many, many more South Australians will know who AIML are, and I have no doubt that the work they are doing now will really light up the sky.

They are doing some amazing work there already and they have some of the world's leading personnel in this field. People often say to me about research and other things, 'Oh, well, South Australia is just good old South Australia,' but you can look at the fact that people who come from our state and who work in our state can do things that can change the world. You only have to look at the person after whom my electorate is named, and there is no reason why some of the things we do in South Australia, and that AIML do in particular, cannot have an equally large impact on the lives of ordinary people.

Motion carried.