The Second Machine Age: Work, Progress and Prosperity In A Time Of Brilliant Technologies

By Myriam Melki, Pam Liou, Sam Lavigne, Jon Wasserman

Erik Brynjolfsson and Andrew Mcafee begin the Second Machine Age with a broad discussion of human progress, asking: “What have been the most important developments in human history?” They argue that human social progress corresponds with technological progress, and that the last great leap in human progress can be directly traced to the invention of the steam engine. Furthermore, we, in our current historical moment, are in the early stage of what will become the next great technological and social breakthrough. Where the previous leap had been brought about by our ability to leverage the power of steam (and fully expressed itself in the industrial revolution), the current leap is catalyzed by advances in computer technology, and will lead to what they call a “second machine age.” And what, exactly, will human progress look like? “We’re heading into an era that won’t just be different; it will be better, because we’ll be able to increase both the variety and the volume of our consumption.”


In chapter 2 Brynjolfsson and Mcafee point out how computers have become increasingly good at performing tasks that were previously assumed to be impossible for a computer to complete. For example, self-driving cars, once thought to be beyond computation, have now become a technical reality and will likely soon enter the consumer space. At the same time, computers like the Jeopardy machine “Watson” are getting better at processing natural language. They argue that three key characteristics are most foundational in the “second machine age”:  technology is becoming increasingly “exponential, digital, and combinatorial.”


Chapter 3 tackles the exponential quality of technology. Brynjolfsson and Mcafee reference Moore’s law and describe how computer technology advances at a consistently exponential rate, due mostly, they claim, to the ingenuity of computer engineers and designers. They argue that exponential advances accelerate so quickly that they are difficult to fully comprehend.


Chapter 4 describes the power of digitization. Brynjolfsson and Mcafe’s main point is that information is becoming increasingly digitized, which increases overall understanding “by making huge amounts of data readily accessible.” Large quantities of cheap data can be analyzed and collated. As an example they cite Waze, a GPS app that collects information about road conditions from everyone who has the app installed, thereby converting smartphones into data collection devices.


Some thoughts: the description of our current state of technology seems accurate. This is, however, a utopian, pro-capitalist, pro-consumerist book that enshrines a new form of exploitation, one that is distributed and crowdsourced. Technological advances will continue to replace human labor. The authors give this march of innovation an uncritical positive value. Unfortunately, the utopian goal to free up human time can never happen because the owners of innovative technology don’t distribute the surplus value they create. It’s difficult to benefit from “exponential, digital, and combinatorial” advancements unless you control the robot that replaces you.


Chapter 5 starts with a very interesting quote, one the whole book seems to be revolving around: “Productivity isn’t everything, but in the long run it is almost everything” (Paul Krugman). The authors emphasize the importance of general purpose technologies. GPTs such as steam and electricity impacted more than just their respective industries, they spread quickly to other sectors of the economy and revolutionized the industrial world. And Information and Communications Technologies are the new GPT. And although innovations get used up, causing the economy to be stagnant at times, recombinant growth generates new ideas, thus boosting the economy once again. All inventions are a mish mash of things invented in the past and overlooked. There are endless possibilities for new ideas because there are always ways to recombine things and ideas in new ways, especially in the internet age. In the early stages of development, growth is constrained by number of potential new ideas, but later on it is constrained only by the ability to process them. The solution would presumably be to bring in more eyeballs in order to process more ideas. According to the authors “Plenty of building blocks are in place, and they’re being recombined in better and better ways all the time”.


Chapter 6 is entitled Artificial and Human Intelligence in the Second Machine Age. The authors seem confident about the future, because “We’re going to see artificial intelligence do more and more, and as this happens costs will go down, outcomes will improve, and our lives will get better. Soon countless pieces of AI will be working on our behalf, often in the background. They’ll help us in areas ranging from trivial to substantive to life changing.” IBM are building the world’s best diagnostician, a robot. And C-Path is a computational pathologist that is supposedly more accurate and less biased than human pathologists. Moreover, the digital network has supposedly led to an overall improvement in all fields, including the environment (air quality for example).


Chapter 7 gives an overview on the important productivity growth that followed the introduction of Information and Communications Technologies in our lives. Despite its “productivity paradox”, this GPT has led to an improvement in various sectors of the economy. The authors then explain how the introduction of the internet and sometimes even just organizational softwares in firms and industries have improved in the long run their productivity.


The Second Machine Age details the shortcomings of the current models of measuring economic growth: GDP. Are there alternative metrics to articulate the productivity and wealth of a nation other than GDP? What does GDP fail to capture?


Free digital goods pose a profound challenge to qualitative measurement because they offer value and improvements to quality of life without driving revenue. These innovations create efficiencies; however, the jury is still out on whether these benefits outweigh the effect of an exponentially increasing number of new, free digital goods flooding the market.


Various alternative approaches include:


Would you rather? Method
The choice between for comparable services over time  Produce mark rather insubstantial qualitative gains, while digital services has improved by leaps and bounds.


Measuring Consumer Surplus-
“If you would happily pay one dollar to read the morning newspaper but instead you get it for free, then you’ve just gained one dollar of consumer surplus.” Measured in money AND time; “rapidly growing consumer surplus from price declines in computers increased economic welfare by about $50 billion each year.”


Gross National Happiness- Bhutanese index to measure quality of life per capita across a wide range of categories.


New Products-  Measuring SKU’s or introduction of new items into the market. Think of it as the product always existing, but was infinitely expensive to produce prior to its invention.


Our global economy has reached an event horizon where the rate of innovation in digital tools and commodities has outpaced the rate of innovation in physical/engineered ones. While digitalization has made marked improvements to the manufacture of goods and production, there are limits to those gains (um–like, physics).  Pushing these limits through globalization and aggressive (but common) Supply Chain Management tactics has led to tenuous infrastructures and complete collapses of entire verticals.


How do you account for an economy where the loftiest, highest creative needs are met while the basic needs are not? Maslow would be reeling. SMA treats digital innovations as the same as physical ones, but they would be weighted differently.


Here’s a perfect illustration of this: Korean parents let baby starve as they play with virtual child.


The disproportionate distribution of wealth that privileges the leader in a category and front-runners of all kinds. That’s pretty much all this chapter says, and supports this claim with lots of statistics.


Technology is not a scalar that benefits a cross section of the population equally.

Chapter 12 is basically about the future. We worry or expect that machines will take over completely, they say. And they probably will with the exciting developments like Google’s autonomous car. But never fear, humans still have the upperhand on deviating from rigid prescribed operations. This is, for now, the thing that robots can’t do better than us. They suggest that you should find an area of industry where people are becoming obsolete, and then figure out the one adjacent space where you can provide human expertise, thus capitalizing on the scarcity. Seems like sound advice.


Chapter 13 makes broad suggestions about future forward policy decisions to prolong the human ability to be productive and desirable to an economy. They assert that we can ”encourage technology to race ahead while ensuring that as few people as possible are left behind. They start off the chapter saying, “With sci-fi technology becoming reality it might seems that radical steps are necessary, but… many recommendations for growth and prosperity found in any “Economy 101″ textbook are the right place to start now and for a while.”  The reason for this is that humans can still manage enough logical work better than machines. This could be interpreted as the obvious way to continue participating in this economic system but NOT because the current economic philosophy is sound.


The way to beat the labor force challenge is to grow the economy. There are a number of ways to do that not least of which is to incorporate more technology into education. That’s good because Education has been a “laggard” compared to other Industries. Therefore if we stop being laggards about learning better and using more technology, then by transitive property we will naturally (guaranteed and obviously) catch up to other industries. Because Education is an Industry and in this economic model is bound to compete with other Industries.


Chapter 14 addresses the long-term strategies and challenges. While admitting that History (not human choice) is “littered with unintended… side effects of well intentioned social and economic policies”, the authors cite Tim O’Reilly in pushing forward with technology’s momentum rather than clinging to the past or present. They suggest that we should let the technologies do their work and just find ways to deal with it. They are “skeptical of efforts to come up with fundamental alternatives to capitalism”. The features of private control over production as opposed to governmental work so well, that they’re even used in China, “which is still officially communist”! Their existence is proof of their success and durability and doesn’t need to be challenged or evaluated.


Most of earners are laborers. As androids take over our jobs, employers will be forced to pay lower wages and ultimately cut jobs. One solution that has been floating around for a few hundred years is Basic Income, a format where everyone is paid to have a minimum standard of living. Endorsed even by Richard Nixon, his last attempt was thwarted by caseworkers and administrators of welfare programs, notorious for their large numbers and unbalanced influence, who were worried about losing their jobs. But the authors are wary of endorsing that solution, because “work” provides more than just money, it provides self-worth among other noble attributes.


Chapter 15 reasserts that we are at an “inflection point”, the precipice of another Industrial Revolution. But with technological advancements, we open ourselves up to accidents and malice of greater magnitude. The internet of things amplifies this further. “There’s a genuine tension between our ability to know more and our ability to prevent others from knowing about us. When information was mostly analog and local, the laws of physics created an automatic zone of privacy.”


So what’s in store for us? Utopian future or dystopian future? The Singularity or Terminator? “It’s wise to never say never, but we still have a long way to go” they say. At the end, it all boils down to Uncle Ben.

4 thoughts on “The Second Machine Age: Work, Progress and Prosperity In A Time Of Brilliant Technologies”

  1. I’m curious, did they say how they calculated: “rapidly growing consumer surplus from price declines in computers increased economic welfare by about $50 billion each year” ?

    Also, I agree that investing in education would better the lives of people, especially since it can train those whose jobs have been replaced and ideally not cost them anything but time to do so.

    I read an article based on these very writers in the MIT Technology Review in August, and the comments below are especially interesting if anyone wants to check it out:

  2. We just wanted to add a brief, note about our evaluation and recommendations of the text:

    Brynjolfsson and Mcaffee’s description of our current technological landscape seems fairly accurate: technological advance is proceeding in an exponential, digital and recombinant fashion. In the very least, their analysis reads true from a silicon valley tech-biz perspective, with the caveat that the exponential quality of technology is not a “hard” rule since Moore’s law seems to rely currently more on human ingenuity rather than the the physical properties of silicon.

    Brynjolfsson and Mcaffee believe that their new paradigm will lead to an explosion of innovation and make the world a better place “because we’ll be able to increase both the variety and the volume of our consumption”, a claim that seems dubious at best, sinister at worst. We don’t have any specific recommendations in response to this work. Instead, we propose the following questions to frame a conversation around alternate paths forward: how can we leverage the exponential, digital, recombinant paradigm for emancipatory ends rather than the continued consolidation of wealth at the top? How does innovation affect surplus value, and how can we shift the unequal distribution of that surplus value? For us these are not questions whose answers necessarily lie in the realm of the technological innovation, they are more likely to be resolved in social/political spheres. That said, we wonder how and in what forms innovation can be détourned.

  3. Detournment, like the Situationists? Innovation turned against itself, and against capitalism (and growth requirements)? Pretty cool table stakes.

    Having read these guys, I keep forgetting their *advice*. I guess they say to try to go alongside technologies and figure out how to service them or do the extra human part. Which seems smart until there’s a plug in to do it for you. I know a hell of a lot of html coders who can’t get jobs in a WordPress universe.

    But I’m more interested in the bigger questions you raise.

    The other thing that stands out is this bit about a turning point or event horizon. I had been thinking about it largely in terms of labor, and of physical “carrying capacity” of the planet. But when you quote here:

    Our global economy has reached an event horizon where the rate of innovation in digital tools and commodities has outpaced the rate of innovation in physical/engineered ones. While digitalization has made marked improvements to the manufacture of goods and production, there are limits to those gains (um–like, physics). Pushing these limits through globalization and aggressive (but common) Supply Chain Management tactics has led to tenuous infrastructures and complete collapses of entire verticals.

    It makes me realize there’s a third crisis to be reckoned with. I just don’t understand this one as well.

  4. Brynjolfsson and Mcaffee believe that their new paradigm will lead to an explosion of innovation and make the world a better place “because we’ll be able to increase both the variety and the volume of our consumption”

    I find the cult of consumption quite hollow. I recently read a few articles (“Capital consumption,” Economist, 2012) about how China needs to ramp up its consumption, because its infrastructure expenditures take up too large a share of its GDP. The running theory there is that infrastructure expenditures only make sense when they cause measurably increased consumption. Really? Having a reliable transportation system is worthwhile only if it gets people to the mall, or compels them to buy cars? A hospital is only worthwhile if it increases spending on medical equipment or procedures?

    I’m also reminded of an assertion a colleague of mine once made, that “in times of cataclysm, it is the small and nimble animals that survive.” Are they advocating that we become small, nimble mammals, constantly innovating to avoid starving or being eaten? Like Doug said, it seems like a sad fate for humanity to eek out a modest living on the dwindling resources and tasks not yet suited for automation. Elon Musk claims to have started SpaceX as a response to at least part of this coming dilemma, his goal being the encouragement of space exploration as a hedge against the inevitable resource limits of our planet.

    Along this line of thinking about resources, one of the rote justifications for capitalism is that it is better than the alternatives; or, more specifically, that capitalism provides a resource allocation mechanism that is better than the alternatives. Imagining a society without capitalism requires that you answer some basic questions: what is produced? how much of it? who gets what is produced, and in what quantities?

    I’ve come to find that people who use this pro capitalist resource allocation argument have done well within the capitalist system, and are blind to its negative effects. Frequently those negative effects are hidden (e.g. foreign labor exploitation, foreign unregulated pollution, destruction of local governance) or are marginalized (e.g. asserting that the poor deserve to be poor, or assuming that the exploited are free to work elsewhere).

    I’m fascinated by the idea of using the instruments of capitalism against itself. For example, the strategy of carbon retirement abuses the carbon trading market to further an environmental agenda. Similarly, the Rolling Jubilee abuses the medical debt market to further a socioeconomic agenda. How can we bend the stock market to enable a different agenda? Can we turn the stock market against itself? Do we even need to? Will the stock market just do itself in? And if so, what opportunities would that cataclysm present?

Leave a Reply