Machine, Platform, Crowd: Harnessing our Digital Future (W W Norton) is the latest book from Andrew McAfee and Erik Brynjolfsson. It follows The Second Machine Age which was reviewed here and gives a more detailed analysis of how the technological revolution is coming along. It lacks any direct discussion of legal services but this is a book on the wider picture of what is happening to the economy that, frankly, everyone should read.
The authors give a perceptive and persuasive account of the major economic changes based on the deployment of digital technology. It is all pretty scary unless you, like both authors, ‘have an optimistic vision for the future’ and believe that ‘the next few decades could and should be better than any other that humans have witnessed so far’. Even Messers McAfee and Brynjolfsson, with all their enthusiasm for a new millennium, have to admit that ‘this is not a prediction: it’s a possibility and a goal’. We all need to work out how we are going to respond to the titanic forces the authors so convincingly describe: as they admit, ’No single future is predetermined. Just as individuals can chart their own courses, so can companies and so can societies’.
The central thesis is that the second machine age is coming in two phases. The first started in the mid-1990s as the deployment of computers began to cause a significant rise in productivity once they had taken over various aspects of routine work. The second phase opens around 2010 illustrated, for example, by Google’s announcement that it has been experimenting with autonomous cars. By 2012, smartphones connected over £1bn people (the iPhone was only introduced in 2007). In this second phase, technology is moving to process tasks which are not routine – such as winning at Go or developing other uses of artificial intelligence. Vast numbers of people have instant and constant communication between themselves and the internet: ‘They can also engage in many kinds of exchanges and transactions, bringing billions more participants into the modern global economy’.
The authors may be at the forefront of modern thinking but history still pulls at their coat tails. They organise their thought around the ancient magic of the number three – falling in line with ancestors like the Greeks and Egyptians. Thus, the modern age, for them, is characterised by three significant shifts in work patterns: from man to machine; products to platforms; and from ‘the core’ to the crowd. People used to design products but now computers have a significant role through their ability to manipulate unfathomable amounts of data. They give the example of how a heat pump was completely re-designed and built by a computer and an attendant 3D printer in a way which was much better than anything a human had done. Such physical products are, of course, still necessary but emerging giant commercial organisations like Uber, Airbnb, Alibaba and Facebook all trade withoutpossessing much in the way of physical assets. The transfer from the certainties of the ‘core’ to the wisdom of crowds is seen in the move from the Encyclopaedia Britannica to Wikipedia or by the emergence of the virtual realities of blockchain and bitcoin.
One insight in the book that is relevant to law is the discussion of human minds can be ‘brilliant but buggy’. Humans may think they are better than machines but, actually, given adequate data, programming and some degree of human supervision, machines are often better. They can, for example, make better diagnosticians than lawyers or doctors. The authors are good at detail and caveat: they are clear about the limitations of computer-based decision-making. Machines tend to be worse than humans in taking everything into account – even that which seems extraneous: ‘they have great difficulty gathering more or different data from what their builders and programmers allowed.’ Rather comfortingly, ‘Another huge advantage that humans have is good old common sense.’ They have some good examples. In the wake of an early terrorist attack, Uber’s automated procedures started to raise prices for rides as demand soared from people wanting to leave the area affected. This was, to put it mildly, not good publicity; required manual override; and subsequent rejigging of the process. But Uber’s management did respond and it should not happen again.
One of the strengths of the book is that it is not all theory. There is extended and very practical discussion both of developments that worked and, as interestingly, those which did not. For example, Airbnb has established itself solidly in the tourist but not so much in the business market. Why? Well, it turns out that business travellers really value location and services – variable characteristics not quite so amenable to bulk transactions as tourist accommodation. A study in Austin, Texas, found that over five years there had been a 10 per cent decline in overall hotel revenue – presumed due to Airbnb and similar platforms – but the impact was located at the cheaper end ‘with lower priced hotels and those not catering to business travellers being the most affected’. Some of the higher end hotels must, of course, be experiencing some effect from other platforms such as Tripadviser or Hotel.com. But neither these nor Airbnb have radically disrupted this part of the hotel industry.
Of the three identified trends, it is probably machines which will most affect the law. You can already see the enormous interest of the large corporate firms in artificial intelligence and new technology – just one example of which would be the joint investment of BakerMcKenzie and Allen and Overy in Ulster University’s new Innovation Hub. You can see also the emergence of platformslike Avvo in the US which seek to match a market of consumers with a market of providers. In terms of decentralised crowd funding, you can see the emergence of entities like CrowdJustice. You realise from looking at the early adopters in other fields, however, just how far the law has – and is likely – to go, particularly for those on low incomes.
And here is a problem for legal services for those on low incomes. How will legal services develop when the financial rewards for success are low, non-existent or even negative (as when more clients mean more work but no more money)? It is unclear whether there will be a trickle down effect from commercial services where AI and machine learning becomes routine. Or will legal services for the poor remain unmodernised, under-funded, dependent on the scattered provision of overworked staff operating without much technological assistance just as a few feudal field systems survived into the 19th century in rural Britain? Will platforms develop where providers compete to deliver low cost services to consumers who may only have five or six legal transactions in their lifetime? Will the pro bono link between large corporate firms and struggling legal service provision provide a unique route for the transmission of technology – which once devised will have relatively little on-cost? Does the crowd have much of a role other than in occasional litigation funding?
This takes us into the territory explored by Richard Susskind in relation to the legal profession but we can also use a reference in his recent writing to take us back to the mainstream. In The Future of the Professions written with his son Daniel, Professor Susskind predicts the decline of the barista and custom-made coffee in favour of automated coffee machines. Well, that is one for the dustbin. I am pretty sure that my area of London – admittedly now emerging from decades of grime and decline to something close to trendiness – is typical in its barista boom. And, personally, I am certainly not paying around £3 a shot for a Nescafe capsule though I happily shell that out at a number of local coffee houses for an individually made cappuccino. And the point? Well, fighting against all the force of technology may be the inherent fact that humans are social animals that crave social interaction. Lawyers, even in big corporate firms, are going to have a role as communicators, advisers and supervisors of machines. There are just not as likely to be so many of them. We need much more discussion of all this.
We also need more discussion of the economic and political power shift that is occurring in the new economy. My search for annual revenues of the top four digital businesses suggests that they were as follows in 2016: Apple $215bn; Amazon $136bn; Google $90bn; Facebook $28bn. That is a combined figure of $469bn – roughly the size of the GDP of Poland, 24th largest economy in the world. That represents enormous undemocratic power and it is not politically neutral. These companies share with those like Airbnb and Uber a fanatical attention to delivery and detail. They apply that to the whole of their business which they understand to include a hostility to public regulation and state taxation. They share what seems to a European a particularly American approach in which they are socially liberal – and many of their major figures supported Hilary Clinton and opposed Trump – with what for us is a pretty right wing undermining of the role of the public and the state. We certainly need much more discussion of that.
So, this book will not give much direct assistance to a legal services lawyer or manager seeking to figure out how technology might impact on their work. But, even if your time is limited, I would read it. We need to understand – and debate – the wider context in which we are working.