Rise of the Robots

By on June 29th, 2017 in Book Reviews, Magazine Articles, Robotics, Societal Impact

Rise of the Robots: Technology and the Threat of a Jobless Future. By Martin Ford. Basic Books, 2015, 224 pp.

Reviewed by Eric Stayton

Machines and mechanization have a long history of displacing human workers by eliminating the need for certain types of human labor. This process has had its casualties, but is generally acknowledged to be the source of long-term economic prosperity. While we may mourn the loss of handicrafts, few would choose to do away with the fruits of the multiplicative forces of mechanization and automation. We owe our standard of living – from affordable goods to accessible public utilities – to the productive potential of machinery; a combination of education, job-retraining, and social services have served to address the plights of those displaced. For much of recent history, our optimism has been borne out, with increased prospects, prosperity, and job opportunities – if not necessarily more fulfilling ones – for the more highly educated.

But this time, could it be different? Does information technology represent a new face on the same old problem? Or do its unique affordances change the situation in a fundamental way, for which history provides neither models nor solutions? This is the question that Martin Ford’s Rise of the Robots seeks to explore in its first half. In this capacity, Ford’s book carries echoes of the arguments of many intellectual forebears who have studied manufacturing technology and factory automation – Hugh Aitken, Norbert Wiener, John Diebold, David Noble, David Nye – though few of these are referenced in his text. Instead, history is not Ford’s focus; and while the past is mined for counter-arguments to be anticipated, Ford’s thesis is that the situation is indeed different this time, and far more dire in terms of the ultimate fate of human labor. The book’s second half examines the intersection of developing technology and social problems. How do we address the coming impacts of technological change?

Rise of the Robots is laudable as a trade book about emerging technologies. It presents a broad, sweeping overview of the issues at hand for the modern laborer in the increasingly automated economy. From manufacturing to the service sector, from elder care to intelligent tutoring, myriad applications of artificial intelligence are addressed – enough that, if one accepts Ford’s claims about the impacts of increasing automation, there is no conceivable alternate economic sector in which those job losses can be offset. This multiple-industry focus distinguishes his book from those focused specifically on manufacturing or commercial robotics, and aligns it more closely with contemporary works like Brynjolfsson and McAffee’s Second Machine Age, though Ford is ultimately less optimistic about the future. Also key to this work’s approach, Ford fuses the technical, including approachable (simplified) descriptions of technologies like artificial neural networks and genetic programming, with economic and political information and critique.

Rise of the Robots uses descriptions of advancing AI technology to set up its punchline: the economy must be restructured to avoid total collapse. To this end, Ford provides a number of insightful and productive re-framings of salient economic issues. Offshoring becomes “virtual immigration,” since remote workers profit from the economy and the infrastructure of a nation without directly contributing to its economy. Similarly, Ford asks if decades of public investment in information technology (via DARPA, NSF, and other organizations) should give citizens a claim on the ownership of society’s “accumulated technological capital” that is otherwise at risk of being captured by a small, elite minority. The economy, in Ford’s eyes, can be likewise reinterpreted as a resource in which all citizens have a stake, and therefore can be used to rationalize a guaranteed minimum income. Ford’s book is strongest as an approachable and compelling entry to these sorts of economic, political, and moral questions, and why they are not mere flights of fancy. We have to be ready to rethink many of our historical assumptions about technology and the economy in order to adapt to increasing automation. This much is clear from Ford’s account.

Ford tells great stories, both about innovation in the last 50 years and about the potential impacts of widespread automation of work in the future. But while they are suggestive, useful vehicles for thinking about the future, the stories do not provide factual concrete proof. This is in a sense a fundamental problem with attempting to study the future, but could be ameliorated by stronger academic citations. Despite its name, Rise of the Robots is less about robots than it is about adapting the economy to current and future problems, and its solution-making is far more interesting than its coverage of the technology itself. Education and health care, as described here, suffer inefficiency more because of economic factors than their relative lack of information technology and automation. Entrenched forces of bureaucracy, combined with the availability of large amounts of money – from students and families in the former case, and insurance companies in the latter, have combined to allow for costs to increase. While MOOCs, medical robots, and machine-learning-powered medical decision assistants may cut down on costs in both sectors, Ford’s suggestions are primarily about social policy. His prescriptions here are convincing; but when they lack references, the work suffers.

The economy, in Ford’s eyes, can be likewise reinterpreted as a resource in which all citizens have a stake, and therefore can be used to rationalize a guaranteed minimum income.

Even if one is relatively willing to accept Ford’s argument that fundamentally high ceilings for artificial intelligence put just about every occupation at risk – including knowledge-work and even “blue sky” innovation, in the extreme case – the details are fuzzy. Ford may speak from significant experience in industry, but does not convey that experience in a way that allows for independent assessment and corroboration. Limited citations often reference contemporary news articles, which make it difficult to assess how much of what is being discussed is real, and how much is hype from the technology companies concerned. He is also optimistic, perhaps overly so, about the potential of “Big Data” to power the next revolution in decision-making and AI applications. While he does recognize some of the perils of statistical decision making based on even very large data sets, he presents certain cases, such as Target’s use of buying patterns to perform “rather creepy” pregnancy prediction, without critique: telling only the successful side of one venture does no justice to the challenges of appropriately and robustly applying correlative patterns on a large scale. Nevertheless, the author’s fundamental points for the future stand, even if one were to quibble about approaches and timelines.

Many such issues with Ford’s book are ameliorated by its far-reaching perspective. Not content to examine only the next 10 years, Ford targets the possible future that current trends point towards. Though he maintains a healthy skepticism about an impending “Singularity” and the feasibility of strong AI in general, he remains that will be available for them.” This last section is where the book truly finds its feet; using all of the previous discussion as background, Ford asks how we prepare for an essentially inevitable situation. It seems likely, as Ford argues, that no amount of new technology and associated profit opportunities – blogs, marketplaces, services – will fill the subsistence gap for most of the citizenry. A major shift in economic policy is the only solution that Ford sees to ensure that consumer spending can continue to drive the market. Short of a complete rethinking of the entire concept of capitalism, which he does not contemplate, a guaranteed minimum income seems to him the best option.

Ford makes a compelling case for this, from a pragmatic point of view. He consults the works of multiple economists, though it was unfortunate to see little engagement with what philosophy has to say, inasmuch as that is relevant for what seems to be a fundamental ideological and societal shift in welfare policy. Nevertheless, the author builds a strong, if not closed, economic case for why we need to start moving toward a system that is robust in the face of the next phase of worker displacement. What lies beyond – how we produce longterm stability in a world where most people are left with no feasible productive work they can do – is left as a subject for another book.

Overall, Rise of the Robots is well-constructed, aimed at a popular audience that knows little about the underlying automation technologies or the social and economic realities of competing with automation. For a technologist, the coverage of the “robots” themselves is not particularly novel or insightful beyond what a consistent reader of Wired, Technology Review, and other technology-focused publications would already be familiar with. And for someone who attends to debates in contemporary economics or public policy, much of the economic and policy information is unlikely to be ground-breaking. But while little of Ford’s book is on the cutting edge of any one sub-field, Rise of the Robots is a competent, approachable, and well-written synthesis of information across many areas, and provides a valuable, coherent picture of automation’s socio-economic interactions. While those looking for a technical deep-dive will likely be disappointed, his well-reasoned perspective makes the book worthwhile for general readers.

Reviewer

Erik L. Stayton is a doctoral student at M.I.T. in the History, Anthropology, and Science, Technology and Society program. His research interests include artificial intelllgence, automation systems, and human-computer interaction. Email: estayton@mit.edu.

 

http://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=7484828