For the first few years after its invention, the laser was described as ‘a solution in search of a problem’. Now lasers are everywhere. They’re used to scan barcodes, remove tumours and analyse chemical compounds. But initially no-one was quite sure what to do with this new technology. We have the opposite problem today. We’re facing down a wall of radical inventions and innovations that we can easily imagine will transform our world.
Take autonomous cars—the most public and obvious change that is now just years, perhaps months away. Autonomous vehicles are already being used across our transport networks. Driverless trucks shift iron ore out of mines. Driverless trains move minerals across the Pilbara. Pilotless cargo ships send goods across the planet. Self-driving vehicles for consumers will change the way we commute, how we travel, how we relate to distance, sprawl and density.
Autonomous vehicles are possible because of advances in a few fundamental under-lying technologies—smart sensors, data mapping, artificial intelligence, machine learning and neural networks. Autonomous vehicles need high-resolution maps of the world around them, so cartographers are building digital maps of the world that are close to a 1:1 scale and dynamically updated. Some autonomous systems teach each other about obstacles and unmapped hazards in real time—the computer in an autonomous system draws its intelligence from the network, not just its own power. Machine learning and neural networks are set to be endemic in every industry, every supply chain, every ‘production function’ (as the economists would say) in the economy.
In the next decade we’re going to see biological and chemical breakthroughs join these advances in computer science. Biological innovations—such as CRISPR gene-editing technology—allow us to tackle disease and human ailments at the most fundamental biological level. When the economic historian Joel Mokyr was in Melbourne in early December, he told his audience that these nondigital innovations and inventions are just as likely to shape our future—in work and as a community—as any of the more prominent digital inventions.
The regulatory and public policy hurdles facing these changes are of course immense. Consider again the challenges posed by autonomous vehicles. Road rules have to be restructured. Infrastructure may have to be redesigned. Figuring out the legal liabilities of vehicles that are in accidents is a huge issue. Who is to blame in an accident: the driver, the company that wrote the autonomous software or the network of other drivers and other autonomous units that mapped the obstacles? How we regulate gene editing, robotic ships, distributed autonomous organisations, cryptocurrencies, 3D printers so powerful that they can print illegal firearms, and so on will be a problem for federal and state parliaments for decades to come.
• • •
New technologies always have distributional consequences. Jobs are replaced or eliminated in some sectors and not others. Some workers find themselves in a bull market, some in a bear market.
In an excellent book, Changing Jobs: The Fair Go in the New Machine Age, Jim Chalmers and Mike Quigley outline from a social democratic perspective how artificial intelligence, automation and robotics might change the industrial relations system, effect the education system and influence patterns of inequality in Australia. Chalmers is the member for Rankin in Queensland and Quigley a former telecommunications executive. Their book represents what is hopefully the start of a parliamentary reckoning with long-term technological trends.
Chalmers and Quigley don’t tell a hackneyed ‘robots will take all of our jobs’ story. They try to reckon with the now inevitable: any job that is repetitive or can simply be represented by an algorithm will very shortly be automated. The jobs in those categories are blue-collar and white-collar. Low-end white-collar jobs such as call centres have already been automated. High-end white-collar jobs such as many legal industry jobs are also likely to be automated.
The first question is, what happens to the people who now perform those roles? This is a problem, but not a new one. We have managed these sorts of structural shifts before—sometimes well, often poorly. A combination of reskilling (both publicly subsidised and privately funded), social welfare investment and (unfortunately) premature retirement is the usual approach.
An equally pressing question is how to prepare new workers for this new age. Chalmers and Quigley rightly put a lot of emphasis on education, and the sorts of education they foresee as necessary for an era of disruption. The key skills the authors identity are the ability to self-educate, formal maths and science education, and proper statistical thinking. The problem with any recommendations about the future of education is it is hard to plan for a future that has never been less certain. Happily Chalmers and Quigley do not insist all Australian students learn to code. This proposal (which incidentally is Labor Party policy) is faddish and short sighted. More people should learn to code, of course. Computer programming is going to be increasingly in demand. But just as everyone who drives a car doesn’t need to know how an internal combustion engine functions, coding will remain subject to specialisation and the division of labour.
The third question Chalmers and Quigley address is inequality. They reject, rightly in our view, the idea of a universal basic income (UBI)—a fixed standard ‘welfare’ payment given to all citizens regardless of their employment status. Though they do not make this argument, the theoretical appeal of a UBI is that it is given to everyone unconditionally and replaces the vast majority of other transfer payments. The political system being what it is, no such theoretically pure policy is ever likely to pass the Australian Parliament, and an imperfect UBI may be worse than no UBI at all.
Rather, Chalmers and Quigley propose a range of less ambitious reforms to the existing social welfare system. For instance, they recommend ‘a “social safety net” that uses big data for good in the social security system’, more emphasis and attention paid to caring roles and bringing people with disabilities into work, and income smoothing for taxation purposes. They are oddly sympathetic to Bill Gates’ idea of a tax on robots. Robots, of course, can’t be taxed—only their owners can. When does a ‘machine’ become a ‘robot’? Are algorithms robots? Nevertheless, their interest in a robot tax represents the limit of their radicalism.
Changing Jobs is a very valuable contribution from a parliament that is hardly awash with deep thinking about the future. We can only hope that some enterprising liberal or conservative politician is thinking about these ideas as well. But to our minds Chalmers and Quigley make a key fundamental error—one made by nearly all of the best thinkers on this topic from Nobel Prize winners down. That is an assumption that the institutional structures of the society will remain fixed while new technologies are squeezed into them.
Consider again the idea of a tax on robots. Gates would like to peg the tax to the salary of the worker that the robot replaced. If the worker was earning $50,000 a year, then the robot tax would be equivalent to the income tax that worker would have paid. The assumed economic dynamic seems to be this: one robot joins the assembly line, one person leaves the assembly line. But the factory remains. Yes, the factory may be relocated to China or Bangladesh. But it remains as a discrete unit of production: four walls, large and expensive equipment, and a single corporate owner.
We don’t think this is how it will be. Along with robots, automation, machine learning, gene editing and neural networks, we are now seeing a revolution in how our economic institutions are structured. This revolution in governance will have profound effects on how we as individuals and communities interact with old and new technologies and institutions. But to explore this we have to talk a little bit about blockchains.
• • •
Blockchains are the underlying technology that powers cryptocurrencies such as bitcoin. A lot of ink has been spilled trying to identify who the bitcoin inventor, the pseudonymous Satoshi Nakamoto, is. But more important than Nakamoto’s personal identity is the community from which he emerged—a group of ‘cypher-punks’ or ‘crypto-anarchists’ who in the 1990s and early 2000s were experimenting with the use of cryptography (as one prominent member, Timothy C. May, declared) ‘fundamentally [to] alter the nature of corporations and of government interference in economic transactions’.
What did Nakamoto invent? Digital currencies are vulnerable to the ‘double spending problem’. Any digital item is easy to copy. If we want to create a digital currency, what stops a holder of a unit of digital currency from copying it and spending it twice? Previous solutions to the digital spending problem relied on having some central authority validate transactions to ensure money wasn’t being spent twice. Nakamoto’s invention was the blockchain—a mix of existing technologies that allowed a distributed ledger of digital currency to be updated securely without any need for a trusted centralised authority.
It turns out that blockchains can do much more than power digital currencies. Block-chain technologies developed in just the last couple of years allow people to write contracts that self-execute, form organisations securely and across national borders, and shift records of ownership and property at close to cost and instantly anywhere in the world.
Blockchains are fundamentally a technology of governance. They are not perfect. Right now blockchains are expensive to run and often risky to use. The history of blockchains starting with bitcoin is, undeniably, a history of scandal, criminal activity, fraud, incompetence, speculation, a fair bit of disappointment and massive uncertainty. But it is not unusual for any new technology, especially one so open to the public, to be targeted by fraudsters and opportunists.
Blockchains are significant because (if nothing else) they are a proof of concept for a form of economic governance that we didn’t know was possible. We know now that it is possible to run a decentralised ledger—a ledger spread across a computer network—without the need for any single central authority in charge. And it turns out that ledgers are everywhere in the economy. After bitcoin we now know that money can be thought of as a ledger of ownership. Indeed, much of what governments do is manage ledgers—ledgers of property titles, ledgers of taxation obligations, ledgers of entitlements, ledgers of citizenship.
But firms are ledgers too. Firms are networks of contracts and capital arranged in a way that produces economic goods. Imagine a firm as a list of relationships that maps who works in what department, who has responsibility for what production, which machines and production inputs are owned (and where to buy more of them), and how primary inputs move through the firm to become useful things to sell to others. That’s a ledger.
Firms are hierarchical because their ledger has to be managed, operated and updated. New economic conditions, changes in the costs of inputs, changes in consumer tastes, changes in the workforce all demand a managerial class to make strategic decisions that can filter down the hierarchy. Alternative corporate forms—such as workers cooperatives—have not thrived at any scale because they have been unable to make the sort of strategic moves at which traditional large firms excel. Blockchains offer a new way to structure a cooperative firm: to achieve decentralised consensus about economic and strategic priorities among workers with a common interest.
We’re used to seeing technological change in production. Electricity, the internet, lasers, penicillin, the aeroplane, mobile phones—all have had huge effects on our lives, but we sort of know how to integrate them into our thinking, even as they rip up industries and certainties as they go.
But we don’t see technological change in governance very often. Arguably the last revolution in governance was the invention of the corporation—the joint stock company of the seventeenth century that became the governing structure for corporate and financial capitalism in the twentieth century. Perhaps we could say that representative democracy (the parliament) is another such structure of governance.
Distributed systems allow production to be distributed too. Those single four-walled factories could be obsolete. Why own expensive capital equipment when you can easily and flexibly rent access to equipment when needed? One interesting blockchain application is Golem: a decentralised, distributed network that allows users to rent idle computing power on any computer signed up to the network anywhere in the world.
The owners of that computing power are paid with Golem’s native cryptocurrency GNT. Since the Second World War firms have been installing supercomputers for computationally intensive tasks; now that sort of investment can be spread globally across thousands of idle, less powerful, less costly computers. And it can be done without the need for a trusted authority or firm to manage the service.
This sort of application is not trivial. Hollywood needs a massive amount of computer power to render complex CGI scenes. Academic researchers need access to powerful computers to exploit the huge volume of data now available. As economic activity becomes digitised—more and more of us now spend our lives producing while sitting in front of an LCD monitor—the possibilities for this sort of simple decentralisation and disaggregation of capital investment grow. The demand for cloud computing is a big factor underpinning the competitive dominance of firms such as Amazon and Google. In these early blockchain experiments, we can see a vision of a future where those large firms compete against open protocols.
The technological revolution we face consists of revolutionary production technologies matched and empowered by revolutionary governance technologies. Mass production is ceding priority to mass customisation. We will order custom products sourced from across the globe by suppliers that are being coordinated not necessarily by people but by artificially intelligent, automatically self-executing production lines.
• • •
Popular writing on the future of work is not exactly blind to changes in economic governance. When we talk about the gig economy, the sharing economy or the increasing casualisation of the workforce, the growth in independent contracting (real or ‘sham’), we are really talking about changes in the structure of the firm, changes in the way we relate economically to each other, to our ‘employers’, and to the disaggregation of the mid twentieth-century big corporate form.
The sharing economy refers to the idea that mobile phone technology can be utilised for short-term use of idle resources (cars, drivers, rooms). It’s controversial for many social democrats in part because it still looks a lot like a variation of the employer–employee relationship. Uber is still a company, Airbnb is a company. But those institutions are now on the cusp of change. Even Uber, the great disrupter, can be disrupted. If you want a vision of the blockchain economy, imagine a decentralised Uber, where drivers and passengers find each other on the street, securely and safely, without the need for a big American company to manage their interaction. That’s what May meant when he talked not just about preventing government from intervening in the economy, but undermining big corporations as well.
In this context, the questions raised by Chalmers and Quigley are even harder to answer. Even high-tech, highly educated, highly skilled workers fully versed in coding are going to be facing an economic landscape that looks completely different from what we have now. Even fundraising and venture capital—the way we finance new projects—will be done in new ways. The ICO craze in the second half of 2017 (an ICO or ‘initial coin offering’ is a way of financing blockchain applications through the sale of the cryptocurrencies that power them) was rife with scams and frauds but nonetheless offered a vision of how even the fundamentals of industrial structure are up for grabs. Learning to code will not offer our children the institutional certainties that our parents or grandparents may have enjoyed.
Governance technologies present their own challenges from the perspective of inequality. Inequality is in part a function of what economists call the ‘superstar effect’. Superstars such as Beyoncé and Mark Zuckerberg fill out the extreme tails of the income distribution spectrum thanks to their global platforms and recognition. Globally decentralised markets powered by distributed networks raise the possibility of superstars in all walks of life. When it is possible to hire the best programmer, accountant, doctor, consultant, lawyer or manager on the planet and integrate them seamlessly into local economic activity, the world’s best are going to enjoy the sort of incomes that were previously reserved for sports stars and musicians. The effect on measures of inequality in this world would be significant.
The blurb of Changing Jobs asks, ‘how should we prepare ourselves, our children and our grandchildren for the changing world of work?’ But before we can prepare we need to understand. Revolutions in governance have their own logic and consequences. Public debate in Australia comes nowhere near these questions—noble exceptions such as Chalmers and Quigley notwithstanding. Long-term reckoning with future trends has been insipid. We can think of the Gillard government’s Asian century, perhaps (if we are being charitable) the Rudd government’s National Broadband Network, and the Howard government’s intergenerational reports. But our political system isn’t even that agile now.
This may come to be a problem. We will inevitably muddle through, but economic transitions are costly and often traumatic. Well-targeted government reform—which will be ceding responsibility as often as assuming it—is not a tool we want to be without. Technological revolutions have made human society richer and better to live in. In the nineteenth century technology pulled us out of the sluggish growth that was until then the natural state of human society. The revolution we have described here is exciting and will make us better off. But we need to be ready for it. •