Journal and Proceedings of the Royal Society of , 2015, vol. 148, nos. 457 & 458, pp. 125-133. ISSN 0035-9173/15/020125-9

Reflecting on the Future of Work in : Pessimism, Optimism and Opportunities

Alan Finkel President, Australian Academy of Technology and Engineering, Melbourne, Australia

e-mail: [email protected]

Abstract This essay has been adapted from a presentation made at the Royal Society of New South Wales’ forum on The Future of Work. The impact of technology and innovation on the future of employment in Australia is examined from both pessimistic and optimistic perspectives. It concludes with a discussion of Australia’s opportunities and suggests a number of measures that could be implemented to help Australia improve its technologically driven prosperity. Received: 28 September 2015 Accepted: 14 October 2015

1. Introduction Which means that the pace of change that we’re all dealing with is beyond that which we Thank you, Donald, for the invitation, and evolved to manage, and is almost enough to Governor, thank you for your inspiration and overwhelm people’s ability to cope. wonderful hospitality. And then, of course, there are the many Today I will assume the role of the techno- predictions of permanent losses of jobs. For optimist, and I will present a broad-ranging example, in 2013 the Oxford Martin School discussion of some of the technologies and at the University of Oxford predicted that in innovations that will impact jobs and job the United States almost half of all jobs in the opportunities. United States will not exist within two decades, lost to computerization (Frey and To start, though, I will put on the pessimist’s Osborne, 2013). hat and share with you some of the many reasons to be gloomy about future Not to be outdone, PricewaterhouseCoopers employment prospects. Then I’ll explain why (PwC) undertook a survey in Australia in I am optimistic, and finally I’ll review a 2015 and reported that 44% of jobs in number of measures that if implemented Australia that we enjoy today will not exist in would help Australia improve its about 15 years from now. They also technologically driven prosperity. predicted that 75% of the replacement jobs will require science and technology skills 2. Pessimism (PwC, 2015).

First of all, if you look at GDP growth, not Hugh Durrant-Whyte and his colleagues at over the last year, but over the last two CEDA recently published a report on the millennia, you’ll see that it’s been flat until impact of computerisation and automation more or less the industrial revolution, when it on employment in Australia. It predicts that started on an exponential rise (Roser, 2015).

125 JOURNAL AND PROCEEDINGS OF THE ROYAL SOCIETY OF NEW SOUTH WALES Finkel – Reflecting on the Future of Work at least 40% of today’s jobs have a 70% studies than we Australians do. We provide probability of being lost to automation in the most of our funding for business R&D next 10 to 15 years (CEDA, 2015). through the indirect means of the R&D tax concession. It is not clear to me based on And not so long ago, in 2011, two MIT personal experience and discussions with economists, Erik Brynjolfsson and Andrew other business leaders that the indirect McAfee, analysed the trends in jobs and investment through the tax system is an productivity and identified a displacement effective utilisation of the money foregone. between productivity and job creation. They predicted that technology will continue to Yet another reason to feel pessimistic is the produce wealth, but not with accompanying stop-start nature of government policy. You new jobs, and therefore there will be a might have noticed; very few policies last for permanent loss of jobs (Brynjolfsson and more than a handful of years. For example, McAfee, 2011). in the innovation sector we’ve had Commercial Ready grants, START grants, If you look at what’s been happening in COMET grants, Commercialisation Australia Australia in the manufacturing sector in the grants, Accelerating Commercialization grants last 30 years, the number of people employed and more. They come, they go; the rate of in manufacturing has fallen from 16% to 8% change is such that none of them are given (Conley, 2014). It’s a staggering change that the opportunity to have long term positive we’ve gone through in just three decades. effect. This is in stark contrast to the United States, where the well-funded Small Business Another thing that adds to my pessimism Innovation Research (SBIR) grant scheme about the future of jobs is the fact that in has been in continuous effect since 1982. Australia our government invests very little to help researchers and other inventors translate Sometimes there are unanticipated drivers their technologies into the market place. A that dissuade innovation. For example, in couple of years ago the OECD published a Australia there is an excellent system for report on competitiveness and innovation. In measuring research excellence. This system, that report they presented the 2011 data for called Excellence in Research for Australia the direct government investment in business (ERA) is administered by the Australian R&D and the indirect investment via tax Research Council (ARC). The ERA collates a incentives. On the direct investment data large range of data to measure research Australia was second last, just ahead of excellence. But because it is heavily weighted Mexico. The leaders such as Russia, the towards publications and citations it United States and South Korea had ten times inadvertently creates a bias within the or more direct government investment than universities against engaging with industry, in Australia (OECD, 2013). because the time that a researcher spends working with industry is time not spent at the A second surprising and telling indication bench or at the computer doing the research from this data set is that the balance between and writing up papers for publication and direct and indirect government investment is consequent citations that will give her completely inverted in Australia. The department a good score in the ERA. It’s not Americans give a lot more direct funding cash the intention of the ERA to discourage to inventors to help them undertake feasibility

126 JOURNAL AND PROCEEDINGS OF THE ROYAL SOCIETY OF NEW SOUTH WALES Finkel – Reflecting on the Future of Work engagement, but in some university undermining their confidence to start departments it has that effect. companies or try new careers.

Here’s another reason to be pessimistic. If But the fact that pessimism is a state of mind the government is not driving innovation, you doesn’t mean that some of the concerns would hope that private firms would step in aren’t real. Take the upcoming driverless and carry the mantle. In some areas they do, cars. Many of the major car companies such but in the so-called high-tech space they as Mercedes, Volkswagen and Volvo are don’t. There is no agreed way to measure working on autonomous vehicles as too are “high-tech” activity, but one way is to look at new entries to the field such as Google. firms that are backed by venture capital. In These new types of vehicles have the the United States and Israel, more than 10% potential to deliver a lot of benefits. They’ll of GDP derives from venture-capital backed reduce the rate of car accidents and deaths. companies (NVCA, 2011; Cohen and Scheer, As we grow older, unlike my 90 year old 2015). In Australia, the contribution to GDP mother who still tries to drive herself despite from such firms is a mere 0.2% (AVCAL, the entreaties of her three children, we’ll all 2013); private industry is clearly not picking have personalised ‘chauffeur driven’ driverless up early stage translation of technology from cars to take us around, which is fantastic. the lab to the market place at an acceptable Further, since the cars will be able to talk to rate. each other and central control computers there will be less congestion. Of course, another reason to be pessimistic is the excess of regulations. We have what I But the impact on jobs will be huge. You refer to as vertical columns and horizontal know about Uber. Uber is a novel model layers of regulations. The vertical columns where private individuals arrive in their cars are the duplications going from federal to to take you to where you need to go. Now state to local government. The horizontal imagine that the cars drive themselves, and layers are the duplications across anticipate a complete change in car ownership departments. This is a well-known problem, patterns so that individuals don’t own cars afflicting many countries. The solutions are anymore. Thus instead of your personal car in some cases to eliminate the duplications. being parked on average 22 out of every 24 In others, to appoint a lead regulator to hours, we’ll have far fewer cars each working coordinate the filings. It is worth noting that a lot harder. But it will mean fewer jobs in the Australian government is in the process manufacturing to make those cars. So the of reducing regulatory duplications. jobs for drivers will disappear and there will be fewer manufacturing jobs. Another driver of pessimism is the media. Nearly every day I read in the paper If anything can justify feeling pessimistic it is something about job losses, or a report on a the onrush of super intelligence. For failure in our tax system, or school education example, in 2011 an IBM computer named that’s not working in Australia, as if we are Watson (after the company founder) won the the most dysfunctional country on the planet. American quiz show named Jeopardy. This is It’s negative thinking. It is rare to see a a quiz show that requires not just knowledge positive report such as “3,000 jobs created”. but intuition and flashes of inspiration. This negative reporting impacts people, Winning Jeopardy was a much tougher

127 JOURNAL AND PROCEEDINGS OF THE ROYAL SOCIETY OF NEW SOUTH WALES Finkel – Reflecting on the Future of Work challenge than winning chess or are dealing with the future shock, even today, backgammon. despite it occurring at a much higher rate than concerned Alvin Toffler. IBM won US$1 million prize money in that game show, but they spent an estimated The classic doomsday prediction that did not US$900 million to $1.8 billion developing come true is peak oil. After the Club of Watson (CNN, 2010). Are they silly? Of Rome published its treatise in 1972, I grew up course not. It was just a means of motivating worrying that I wouldn’t be able to drive a car their engineers, it was a challenge. But now in my later years because there would be no they’re rolling Watson out into very high end oil. The concern started with Marion jobs. It is hard to think of a higher end job Hubbert, who predicted that oil production in than a medical specialist. Well, at the the world would peak in the late 1990s, Memorial Sloan Kettering Cancer Centre in maybe as far out as 2000, and then go into New York, IBM Watson, the computer, acts terminal decline. as an expert computer oncologist providing second opinions to the expert human Well, the reality is, if you look at a graph of oncologists. global oil production it is continuing to rise steeply, with no hint of reaching a peak or In banking, the risk assessors and financial entering a decline. Oil production has grown advisors are on the way out, because IBM steadily right past the year 2000 where Watson is stepping in at the ANZ Bank and Hubbert and others said we should have hit other banks around Australia. These are high peak oil. Why does it keep going up and up end intellectually demanding jobs. and up? Because of innovation, because of technology. Technology can help us recover I’ve given you many reasons to be pessimistic. from the problems that technology causes. Time to flip, and talk about some of the Innovation, or human ingenuity, is always reasons to be optimistic. there in abundance, wiping out the predictions of doom. 3. Optimism Let’s talk about jobs. In 1779, the legendary I’m optimistic because the reality is that Ned Ludd used a sledgehammer to smash the pessimistic predictions are rarely correct. You stocking machines that were taking away jobs hear about impending doomsday all the time, from people in the garment industry in but fortunately we’re still here. Doomsday . Hence the term Luddite. Ned and predictions rarely eventuate. In 1970, Alvin his co-workers didn’t stop the machines. But Toffler wrote a book named Future Shock neither did England enter an era of (Toffler, 1970). In it he talked about the permanent unemployment. stress that people were feeling from the rapid rate of technological change. By today’s John Maynard Keynes, the world famous standard the rate of change was modest. economist, in 1930 wrote that the rate the Video tape was being introduced for United States was finding ways to replace the recording movies and home videos, and the use of labour was outrunning the pace at first mobile telephones were available in the which new jobs could be found (Keynes, form of in-vehicle installations. Well, 40 years 1933). He talked about a new disease – later the rate of change is far, far greater, and technological unemployment. He was you know, what? We’re all coping well. We 128 JOURNAL AND PROCEEDINGS OF THE ROYAL SOCIETY OF NEW SOUTH WALES Finkel – Reflecting on the Future of Work genuinely worried but he was smart enough fail to anticipate major improvements to our to realize that the dislocations to employment health and wealth. were likely to be a temporary phase of maladjustment. Employment recovered. In The biggest change in job sector employment 1961, Time magazine ran a story in which that you’ll ever see is in the farming industry. they concluded that the rise in unemployment In America, 90% of people in the year 1800 in America at that time was due to were employed in the farm sector. I don’t automation (Time, 1961). Fear of automation mean they were necessarily farmers, some got so bad that in 1964 President Johnson were but the rest were rural merchants, or the created a special commission to look into manufacturers of the ploughs or the people what could be done to save jobs in America storing the grain. Employment in the farm from the threat of technology and sector is now down to around 1.7% (Perry, automation. But by 1966 when the 2011). At the same time, the output of the commission delivered its report, the United farm sector has grown enormously. Of States was again approaching full course, as employment in that sector employment. collapsed from 90% to less than 2%, unemployment did not increase by 88%. Despite the fact that new jobs have replaced Instead, new jobs were created. lost jobs again and again, it is always the case that the prognosticators say, “This time it will I’m also optimistic because here in Australia be different”. we have some fantastic examples of innovation. People think of innovation So let me go back to the MIT economists, mostly as being in the high tech sector, in Brynjolfsson and McAfee, who published the venture capital backed companies. But some book Race Against the Machine (2011). They of the big companies do a great job too. In predicted that this time it would be different, high tech, we’ve got the cochlear ear implant, that we are entering a period of permanently it’s done wonderfully well for Australia. But increasing unemployment. our traditional big banks are innovative, too. Our banks have spent and continue to spend Let’s see how good they were at predicting billions and billions of dollars on developing the future, even in the near term. At the time software for their internal systems, and also that they published, unemployment in the software for their customers. Our banking United States was 9%. Today, August 2015, system is, if not the most advanced in the it is 5.1% (Bureau of Labor Statistics, 2015). world, one of the most advanced in the It went down, not up. So in just four years world. Innovative technology serves our they got it completely wrong. Prediction is banks brilliantly and they are making huge difficult, especially about the future, and profits. experts get it wrong all the time. They fail to take into account innovation, because as One of our biggest mining companies, Rio economists they only want to consider things Tinto, has a Mines of the Future program in they can quantify but in the case of which they deploy driverless trucks in the innovation they don’t know how to quantify mines and remotely operate underground it. Innovation leads to the creation and the drilling machines. Everything is either invention of new opportunities. Time and autonomous or controlled by experts back in time again, predictions that ignore innovation Perth. This increases the safety profile of the

129 JOURNAL AND PROCEEDINGS OF THE ROYAL SOCIETY OF NEW SOUTH WALES Finkel – Reflecting on the Future of Work mine, improves the economics and reduces the recent replacement of the Prime Minister the environmental damage. This innovative there might be an even more welcoming approach has allowed Rio Tinto to continue acceptance of the need to do things mining profitably despite the collapse in iron differently. ore prices. So I’ve gone through why I’m pessimistic, I’m also optimistic because I think we do a and why I’m optimistic, but it’s not enough. good job on workforce discipline training in We have to think about where the Australia, and we are getting better. Further, opportunities lie and what we can do to drive at our universities we’re now teaching growth and prosperity. entrepreneurship, we’re setting up mentoring programs and we are teaching postgraduate 4. Opportunities students broader skills in communications and project management. Universities generate research outcomes. Companies want to commercialize proven I’m optimistic because new companies create technologies. We’ve got both these ends of new jobs. This is well documented in the the innovation spectrum in Australia. What United Kingdom where it has been shown we don’t have is a means of getting the that companies created between 2007 and research outcomes from the university to the 2010 contributed 36% of the new jobs in that point of being a proven technology. This is period (Anyadike-Danes et al., 2013). Similar often referred to as bridging the Valley of results have been shown in Australia. Death. As a result, we don’t have enough innovative small firms in Australia. It is important that job destruction should not be considered to be synonymous with There are many things that can be done to try permanent loss of jobs. What the UK study to address that. To start, we can learn from found is that in a 12 year period, from 1998 the United States. There the government to 2010, each year about 28% of the jobs in helps to provide the funds to build the bridge the private sector were destroyed and over the Valley of Death, through programs replaced. Often that’s within the one such as SBIR, many programs run by the company but either way, it is a lot of job Defence Department and programs for churn. We tend to focus on the losses, but medical technology development run by the there’s simultaneous creation. National Institutes for Health. These highly effective programs are why the direct I’m optimistic because we have improved the investment by government in business R&D public discourse in this country about how is so high in the United States, as I mentioned science and technology and research can earlier. We would do well in Australia if we contribute. We’ve got a national science mimicked these programs. policy underway and some research priorities established. The government is talking the Now, of course, in the United States they’ve right talk, although there’s very little action also got venture capital funding and even yet. There are a lot of reviews underway on debt funding for startups. There’s a bank research funding, research training and called the Silicon Valley Bank that, unlike our research infrastructure to support the banks, seeks to invest in high risk companies. research endeavour. I would hope that with We would also do well in Australia if we could mimic these funding sources. 130 JOURNAL AND PROCEEDINGS OF THE ROYAL SOCIETY OF NEW SOUTH WALES Finkel – Reflecting on the Future of Work

necessary skill. Even if you’re trained in arts, We need to attribute value to the effort by you can become a business person, or if university researchers to engage with their you’re trained in engineering you can become industry counterparts. Earlier I talked about an HR manager, because the ability to deeply ERA, which despite being a robust measure analyse is a skill that is required in the of research quality inadvertently has a workforce and is learned by intensively negative influence on how researchers engage studying any discipline. For all these reasons with industry. Cognisant of this, the we need to support training and retraining for Academy of Technology and Engineering displaced workers. (ATSE), of which I am President, has proposed a new metric to encourage Further, we will need new regulatory collaboration between researchers and frameworks to manage the encroachment of industries. It will be a pragmatic measure and artificial intelligence and robots. Artificial it will be quantitative. It will not replace the intelligence and robotics have the potential to ERA, instead it will sit alongside it. do a lot of specific harm. For example, much concern is already being raised about AKMs, What improvements do we need to consider autonomous killing machines. These are to prepare the workforce for the future? One drones that aren’t controlled by somebody often hears that we need more science, sitting in Los Angeles, but instead are given a technology, engineering and mathematics mission, and given the authority to (STEM) graduates. However, in a recently autonomously make attack decisions published paper, Bob Birrell (2015) from according to their mission profile. Monash University analysed demand and concluded that with current government We need to build confidence across our policies we are actually training too many community in starting new businesses. STEM graduates for the number of jobs Returning to the UK, it’s stunning to learn available. It’s sad to hear that, because like that 200,000 firms are born every year many others I have personally been doing my creating about 1 million new jobs. Within 10 best to encourage young people into the years four out of five have closed. However, pipeline. the small fraction of firms that remain will still employ around half a million people We have to train people to work in a (Anyadike-Danes et al., 2013). And all the fragmented workplace of the future where time there are more and more startups that there will be many more people self- make a crucially important contribution to employed, some working in micro or mini jobs and the economy. entrepreneurial start-ups. But at the same time, we must not neglect the need to train So what if my optimism is misplaced? If I’m people to work in large firms. Employees wrong then we need to be planning for a need to be flexible, they need to have deep different society, and now is the time to be discipline knowledge. There’s a tendency to planning. We’ve got to be planning believe that with so much easily accessible strategically. Unfortunately, for quite a few information available it is not necessary to years now there’s been a lack of long term train people in deep discipline knowledge. strategic planning in our national government. That is not correct, because coming to grips We cannot afford to continue that way. with the intricate details of a subject is a

131 JOURNAL AND PROCEEDINGS OF THE ROYAL SOCIETY OF NEW SOUTH WALES Finkel – Reflecting on the Future of Work

It’s not necessarily all bad even if I am wrong that’s the nature of innovation, but it can also and jobs disappear. Remember that loss of create jobs”. We have to optimise that result. jobs does not necessarily equate to loss of wealth. What Brynjolfsson and McAfee from We have to invest in workforce training and MIT were saying is that there will be massive re-training because the job opportunities will productivity increases due to automation, be changing at a rapid rate. And as I said, if I huge amounts of wealth will be generated, but am wrong in my optimistic view of there won’t be jobs. They’ve been wrong employment, we have to start planning a about employment in the last four years, but different society. what if they’re right in the long term? There will be tons of wealth. The challenge is to 6. Discussion/Q&A adapt our society to one in which the wealth can be distributed in an equitable fashion to “Many thanks for your very interesting and the people who aren’t working. stimulating presentation. It strikes me that most of the data and analysis are taken from But wealth without jobs is not a comfortable the USA, Europe and Australia. They focus solution because most people define their on national labour market trends, especially self-worth through their jobs. If our society with regard to cost-benefit considerations goes this way at the very least we will create concerning automation of various types of jobs for psychologists who will be needed to work. Today, much industrial production is help people create their personalized structured by an international division of definition of self-worth outside of labour and by global value-chains. employment. Developed economies still focus on industrial products, even though the actual production Perhaps we can learn how to do this from the work is often carried out in low-wage science fiction literature. One of my favourite economies. In other words a product like a books is called The City and the Stars, by computer or mobile phone may include Arthur C. Clarke (1956). I won’t take you components made in several different through the whole narrative, but wealth countries – such as China, Vietnam and without jobs is exactly the theme that Arthur South Korea – while design and marketing C. Clarke back in 1956 was tackling in this remain mainly in the USA, Germany or quite beautiful short novel, where no one, not another apparently post-industrial economy. a single person, had a conventional working This enables transnational corporations – still job, because everything was taken care of by mainly based in developed economies – to the central computer and the robots. capture most of the value of the product and to maintain high rates of profits, due to the low-levels of wages and social rights in the 5. Conclusion places where most of the work is actually To summarize, we have to accept that the done. Please comment on this observation, pace of change is rapid and will become even and on the prospects for automation in more rapid. Amazingly, young people, economies which today have very low wage middle aged people, all of us, can cope. We rates.” have to invest in innovation to create the jobs - Professor Stephen Castles, Research Chair that replace the jobs that innovation destroys. in Sociology, The University of Sydney. We can’t just say “Innovation destroys jobs” and leave it at that. We’ve got to say “Okay, 132 JOURNAL AND PROCEEDINGS OF THE ROYAL SOCIETY OF NEW SOUTH WALES Finkel – Reflecting on the Future of Work

“It is true that the supply chain is global. For Bureau of Labor Statistics (2015) Regional and State many products that has been the case for a Employment and Unemployment Summary, long time. For example, I made scientific http://www.bls.gov/news.release/laus.nr0.htm instruments in California during the 1980s CEDA (2015) Australia's future workforce?, The and a lot of our components came from Committee for Economic Development of Australia, Melbourne. Japan and various European companies. But Clarke, A. C. (1956) The City and the Stars. accepting that globalization of the supply CNN (2010) 5 billion-dollar tech gambles: Watson, chain is increasing, it is an opportunity as well Cable News Network, as a jobs threat. The opportunity is for our http://money.cnn.com/galleries/2010/technolo domestic suppliers to provide components to gy/1008/gallery.biggest_tech_gambles/3.html global companies. If we don’t do that, and if Cohen, T. & Scheer, S. (2015) Israel's high-tech boom we don’t manufacture much here, the is double-edged sword, Thomson Reuters, manufacturing job numbers will go down http://www.reuters.com/article/2015/01/19/us- israel-tech-m-a-insight-idUSKBN0KS0V520150119 again, starting from what is already a single Conley, T. (2014) US and Australian Employment digit percentage. But jobs will likely increase Trends and the Future of Jobs?, in other sectors. For instance, through my http://tomjconley.blogspot.com.au/2014/01/us company Cosmos Magazine we publish -and-australian-employment-trends-and.html secondary school science lessons. In the old Frey, C. B. & Osborne, M. A. (2013) The future of days, these would have been printed, and employment: how susceptible are jobs to computerisation, there would have been manufacturing jobs University of Oxford (Oxford, UK). for printers, storemen and packers. Today, Keynes, J. M. (1933) "Economic possibilities for we distribute the lessons out of the digital our grandchildren (1930)", Essays in persuasion, cloud, but we employ a team of software 358-73. NVCA (2011) Venture Impact: The Economic engineers to develop and maintain the digital Importance of the Venture Capital-Backed Companies to delivery platform. These are new jobs the U.S. Economy (Edition 6.0), National Venture replacing the manufacturing jobs.” Capital Association - Dr Alan Finkel, President ATSE. OECD (2013) "R and D tax incentives", OECD References Science, Technology and Industry Scoreboard 2013: Innovation for Growth, OECD Publishing, Paris. Anyadike-Danes, M., Hart, M. & Du, J. (2013) Perry, M. J. (2011) Creative Destruction of Jobs Makes Firm Dynamics and Job Creation in the UK: Taking Us Richer, Stock and Developing New Perspectives, Enterprise http://mjperry.blogspot.com.au/2011/07/creati Research Council (Birmingham, UK). ve-destruction-of-jobs-makes-us.html AVCAL (2013) VC Factsheet, The Australian PwC (2015) A smart move, Private Equity & Venture Capital Association PricewaterhouseCoopers Australia. Birrell, B. (2015) Too few or perhaps too many Roser, M. (2015) GDP Growth over the Very Long STEM graduates Australian Universities' Review, Run, OurWorldInData.org, 57(2). http://ourworldindata.org/data/growth-and- Brynjolfsson, E. & McAfee, A. (2011) Race Against distribution-of-prosperity/gdp-growth-over-the- The Machine: How the Digital Revolution is very-long-run/ Accelerating Innovation, Driving Productivity, and Time (1961) The Automation Jobless; Time. Irreversibly Transforming Employment and the Economy, Toffler, A. (1970) Future Shock, Random House. MIT Center for Digital Business.

Dr Alan Finkel is Chancellor, Monash University and from 2016, the Chief Scientist for Australia.

133