The Macroeconomic Impact of Automation
By Alberto Moel, Vice President Strategy and Partnerships, Veo Robotics
Welcome back, dear reader, to our series on the economics of automation. Our previous postings made the case that full-on automation is inflexible and fragile, high-levels of automation can be terribly uneconomic, and an all-human manufacturing approach is suboptimal under many reasonable conditions—the best outcome is a mix of humans and machines working together.
Our work to date has focused on the microeconomics of automation—a worthy topic, and of intense interest to people who actually build factories (and people that pay for them). But policy makers, politicians, talking heads, pundits, and the ever-curious general public care much more about the macroeconomic aspects of automation. Will robots take our jobs? Will robots give us jobs? Will they play give-and-take with our jobs, like a cat playing with a mouse before it does what cats do with mice?
There is a tremendous amount of talk out there on this topic, some of it thoughtful and insightful, much of it echo chamber, alarmist, or badly-premised. As automation, algorithms, and computing power (represented in breathless shorthand as “AI”) continue to evolve, the potential implications on employment, culture, economics, and society in general are now a source of constant fodder for all commentators.
Starting with this one, my next few posts will cover the debate’s many angles, from representations in popular culture and literature, to more focused and sober studies of the macroeconomic impacts of advances in automation. Let’s first take a wide and light-touch look at how robotics, automation, and, more generally, AI have been viewed in pop culture—my favorite by far, and worthy of a special mention, is Rosie the Robot,1 from the 1960s TV series The Jetsons.
Automation as harbinger of violent dystopia
Depictions of robotics, automation, and AI have taken many forms, from the diffuse (2001: A Space Odyssey's HAL 9000 or The Terminator's Skynet) to highly-specific human-like creatures created by humans in our own image. These human-made humanoid forms have a long history in popular imagination, starting in 1818 with Mary Shelley's Frankenstein,2 though Frankenstein's monster wasn't really a "robot" as he was organic and not mechanical.
The concept of a "robot" as a human-made "machine" probably dates back to the creation of the very word robot by Czech playwright Karel Čapek in 1920 in his science fiction play R.U.R.3 R.U.R. is a tale about a factory that makes organic artificial people, called roboti, that can think. In the play, the world soon becomes dependent upon the wealth created by their easily exploited labor. This dependency, combined with world’s blindness to the longer-term consequences of populating the globe with a new lifeform, leads to disaster:
Domin [General Manager of the R.U.R. robot factory]: But in ten years Rossum's Universal Robots will produce so much corn, so much cloth, so much everything, that things will be practically without price. There will be no poverty. All work will be done by living machines. Everybody will be free from worry and liberated from the degradation of labor. Everybody will live only to perfect himself.
The end game (spoiler alert) is that the roboti spin themselves up into a rebellious lather and annihilate their creators and the rest of the human race.
It seems the fear of robots as creepy usurpers bent on the destruction of their creators started at right about the same time as the invention of the very concept of robots. Since then, we've seen countless instances of robotic creatures in plays, films, books, TV series, and other forms of human artistic expression—much of it with a heavy dystopian bent.
A number of contemporary intellectuals have warned of the pitfalls of the development of these “AI” technologies and their (inevitable?) eventual acquisition of super-human intelligence. Nicholas Carr4 brings up the issue of automation and the internet making us (even) dumber than we already are, and the doomsday scenario of machines getting far too smart and doing more harm than good. For example, military and police drones picking their own targets immediately brings up the RoboCop scenario, a dystopia where machines (and their human enablers) bring out the worst in us.
And of course, there's always the Singularity endgame, which goes back all the way to Turing,5 who suggested that a learning machine might achieve a "supercritical" state where, in analogy to the atomic pile, it would produce more ideas than those with which it had been fed.6 In the singularity, a machine would be capable of designing a better version of itself, starting a runaway cycle where smart machines build successive generations of yet more powerful machines, outstripping human intelligence and the ability of humans to control these machines. I think this is just alarmist nonsense, plain and simple.
Automation as harbinger of human irrelevance
Until now, intelligence has always been connected to consciousness. Only conscious and intelligent beings could play chess or drive cars (or, for that matter, invent the game of chess or design cars). But the trend in computation and “AI” is toward the development of unconscious, but intelligent, agents that can perform these (and many other) tasks far better than humans can. Intelligence and consciousness are being decoupled, and if intelligent machines can replace humans, humans will become less valuable.
For example, training a human oncologist takes decades of education and experience. However, when the AI embedded in IBM's Watson engine is eventually trained to be a near-world-class oncologist, it will be trivial to replicate that instance of Watson around the world, providing virtual oncologists to everybody, at all times. The value of human oncologists will drop, except for those "super oncologists" tasked with pushing the boundaries of medical knowledge and training Watson with the latest developments in the field. Whether these unconscious but intelligent algorithms can somehow acquire “consciousness” on their way to artificial general intelligence (AGI) remains to be seen, but I (and many others) remain highly skeptical.
More broadly, automation (in all its forms) is a physical technology, like stone tools, automobiles, or computers, as opposed to a social technology, like money, the rule of law, or social hierarchies. As Eric Beinhocker, the executive director of the Institute of New Economic Thinking at Oxford University and the author of The Origin of Wealth: The Radical Remaking of Economics and What it Means for Business and Society, put it in an interview:7
Physical technologies evolve at the pace of science–fast and getting exponentially faster, while social technologies evolve at the pace at which humans can change–much slower. While physical technology change creates new marvels, new gadgets, better medicine, social technology change often creates huge social stresses and turmoil, like the Arab Spring countries trying to go from tribal autocracies to rule of law democracies. Also, our physical technologies can get way ahead of the ability of our social technologies to manage them – nuclear proliferation, bioterrorism, cybercrime–some of which is happening around us right now.
In other words, automation technologies, conscious or not, are moving faster than we can control or harness them, and with unknown, but possibly detrimental, consequences.
Other commentators, such as Evgeny Morozov,8 raise concerns about privacy and the reliance on technology to "solve" the world's problems. And these themes show up across popular culture—when robots do everything for us, what will we become?
Kurt Vonnegut's first novel, Player Piano, first published in 1952, presents a topical (and perhaps prescient) take. Prepare for more spoilers here—it depicts a post-apocalyptic dystopia where a global war leads to a massive shortage of labor and, therefore, the development of fully-automated, vertically-integrated, and geographically concentrated manufacturing sites. Everything people need is made in these factories, which are supervised by the Engineer and Manager classes.
When the war ends, those out in the battlefields return to a world where their labor is not needed, and they are confined to sites that are like ghettos or refugee camps. One such site is the “Homestead," across the river from “Ilium Works,” an integrated factory site:9
There were a few men in Homestead–like the bartender, the police and firemen, professional athletes, cab drivers, especially skilled artisans–who hadn't been displaced by machines. They lived among those who had been displaced but they were aloof and often rude and overbearing with the mass. They felt a camaraderie with the engineers and managers across the river, a feeling that wasn't, incidentally, reciprocated. The general feeling across the river was that these persons weren't too bright to be replaced by machines, they were simply in activities where machines weren't economical.
The novel also deals with the concerns of the Engineer and Manager classes when "thinking machines" become more proficient and begin to threaten their grip on power:
"…the third one has been going on for some time, if you mean thinking machines. That would be the third revolution, I guess—machines that devalue human thinking. Some of the big computers like EPICAC do that all right, in specialized fields." "Uh-huh," said Katharine thoughtfully. She rattled a pencil between her teeth. "First the muscle work, then the routine work, then, maybe, the real brainwork."
But more pressing is the brewing rebellion at Homestead, where the lack of work is leading to social unrest. This isn't due to poverty or deprivation, as the machines provide food and material goods in abundance for everybody; it’s due to an absence of meaning and direction:
These kids in the Army now, that's just a place to keep 'em off the streets and out of trouble, because there isn't anything else to do with them. And the only chance they'll ever get to be anybody is if there's a war. That's the only chance in the world they've got of showing anybody they lived and died, and for something, by God. You know, used to be you could go out to sea on a big clipper ship or a fishing ship and be a big hero in a storm. Or maybe you could be a pioneer and go out west and lead the people and make trails […].
Vonnegut deals with some of the questions we’re struggling with today: What happens when we don't have to work, but we still want to? Is that a recipe for social unrest?
Automation and our future
The volume of commentary on the implications of these new technologies on the future of work is growing probably faster than the technologies themselves are advancing. Every day brings forth one or more articles in the popular press and in academic journals, and every week one new book tackles the topic.10
We at Veo have some ideas of our own, but as a first step let’s synthesize and translate what's out there. I see four visions in the literature and commentary, which can be conveniently summarized in a 2x2 matrix:11
- The Techno-optimists expect robotics, automation, and AI to continue their inexorable march, leading to increased levels of automation that will generate employment and productivity growth—a bright future. Techno-optimists believe machines will replace humans, but also that the displaced humans will find alternate and meaningful employment as a result.
- The Techno-pessimists predict that this whole brouhaha is likely over-hyped and that we're just going to see "more of the same": continued weak employment growth without increased labor productivity. Techno-pessimists acknowledge the advance of robots, automation, and AI, but do not expect the impact on labor productivity or employment to be meaningful enough to make a difference.
- The Traditionalists also believe that the general applicability of new automation technologies is over-hyped, but also that it will have material impact in some specific applications and verticals, leading to mass unemployment in those areas. Further, traditionalists do not believe these displaced workers will be able to be retrained or find alternative employment. But they expect the transition to be slow, so although dislocations will occur, they could be managed over long periods of time and with good policies, as they were in the First and Second Industrial revolutions.
- The Dystopians, those buzz-killers, expect generalized AI penetration and rapid productivity growth, resulting in persistent mass unemployment and a permanent underclass à la R.U.R. or Player Piano. Bummer.
Kind reader, if you’ve made it this far, I commend you. I’ll stop for now and we will take up the taxonomies of the four visions and what we think they mean in subsequent posts.
1 A sly wordplay on WWII's Rosie the Riveter, Rosie the Robot was a persnickety, error-prone, authoritarian robotic household maid purchased secondhand by Jane Jetson from a shady dealer of old hardware. You can see Rosie’s first appearance in this clip. I grew up watching The Jetsons in Spanish; the show was called Los Supersónicos and Rosie was called Robotina—a stroke of genius in translation, we all have to agree.
2 Some contemporary historians (Feldman and Wilder) believe "Frankenstein" is more correctly pronounced "Fronkensteen."
3 The word robot comes from the Czech word for a form of indentured work, robota, and first appeared in R.U.R. R.U.R. stands for Rosumovi Univerzální Roboti (Rossum’s Universal Robots).
4 The Glass Cage: Automation and Us, by Nicholas Carr.
5 A man ahead of his time along many dimensions.
6 Turing, A., Computing Machinery and Intelligence, Mind 59, 1950.
7 Quoted in Thomas Friedman's opus, Thank You for Being Late. A bestseller by all accounts, but not really my cup of tea.
8 The Net Delusion: The Dark Side of Internet Freedom, and To Save Everything, Click Here, both by Evgeny Morozov.
9 "Ilium" is a fictional town in eastern New York State, which pops up in a number of Vonnegut novels. It's probably a stand-in for Schenectady or Troy, NY. Not coincidentally, General Electric was for many years headquartered in Schenectady, NY. It is now headquartered in Boston, MA, of all places. Full circle.
10 The irony is not lost on us that this very note is classic "pot calling the kettle black," or as they say in Spanish, "el burro hablando de orejas."
11 As a recovering management consultant, it’s hard to break the habit.