[Originally written in July 2018, based on notes and a discussion with Professor Patrick Winston, MIT CSAIL / MIT SLOAN, who passed away exactly one year after this in July 2019.]
Extensive debates around Universal basic income (UBI) have resurged in the context of imminent job losses attributable to Artificial Intelligence and Robotics. UBI has garnered media interest and support by proponents Elon Musk, Richard Branson, Barack Obama, Sam Altman, Mark Zuckerberg and many more. Recent reports point out that 48% of Americans now support UBI.
UBI is a steady monthly income that every adult receives from the government without the need to fulfill a work requirement. It is also known as Negative Income Tax, Basic Income, or Guaranteed Government Income. It’s purpose is to fix or ease the income inequality.
UBI is not a new concept
An equivalent of a UBI was first instituted in 1795 in Speenhamland, England after the industrial and agricultural progress started to take away jobs, as a way to fight property. It lasted for 39 years. In fact, basic income anti-poverty concepts similar to UBI have resurged following job losses after every industrial revolution.
More recently UBI was tried in Finland where it lasted two years. Whether Universal Basic Income failed in Finland or Finland failed Universal Basic Income, the idea of UBI continues to be piloted and experimented. US, Canada, The Netherlands, Iran, Scotland are involved with UBI in varying extents. In the US myriad of programs are underway. Stanford set up a Stanford Basic Income lab in February 2017 to work on government income. Startup incubator Y-Combinator in 2017 funded a pilot program in Oakland, California. Earlier this month, Stockton California decided to give 100 residents $500 a month for 18 months based on a program invested in by Facebook co-founder Chris Hughes. The City of Chicago is also piloting a guaranteed government income program with a $500 monthly program for 1000 people and has its fair share of opponents. The $60 billion Alaska Permanent Fund has been collecting revenue from oil and mineral resources since 1976 to fund a state-wide basic income fund.
UBI may not happen, or at least as we know it
Modern UBI pilots are just starting. No one knows if, how, and when it will work. Both opponents and proponents of UBI have excellent and debatable points and yet to achieve a consensus. It may be a while before economists, politicians, and entrepreneurial billionaires agree on a possible US-wide program.
The political nuances around Universal Healthcare are enough for me to suggest that we do no rely on the government for Universal Basic Income.
UBI will not solve the entire job-loss problem. It will help those who cannot be employed. We have to create jobs anyway.
If UBI does not happen because of lack of consensus, politics, delays, funding or any reason, we will need to create additional jobs that are relevant to the era.
We still have to create new jobs
We have lost jobs in every Industrial revolution but we have also created new era-specific jobs. At least in the past industrial revolutions or “Automation eras“.
- The Industrial Era saw job loss in agriculture but brought millions of jobs in chemical manufacturing, industries using steam and water power, tools development, and factory-based mass-production and mechanization.
- The Technological Era brought jobs in industries that started using electric and oil power. Innovations like railroads, assembly-line manufacturing, telephone, light bulb, phonograph, and the internal combustion engine increased efficiencies and created jobs.
- The Digital Era brought industrial computerization and personal computing which paved a way to enormous efficiencies and productivity, creating many digital jobs. As the internet, Web, cloud, social, mobile became mainstream numerous new jobs were created.
- The Fourth Industrial Era or Digital Era 2.0 is expanding the ongoing digital revolution in the fields of biotechnology, nanotechnology, Internet of Things (IoT), 3D printing, and Internet of systems. Advances in robotics, artificial intelligence, and quantum computing will change jobs as we know them. Autonomous Systems are different because they can make decisions based on what they perceive.
I am calling this the “Autonomous Era”.
Is the Autonomous Era different?
The Autonomous Era brings a new set of questions. Specifically, how will the jobs scene change? So if UBI doesn’t pan out, will we be fine? Will we create jobs as we have done in the past or is this autonomous era prove to be different?
Autonomous Era systems are different from automation (use of automatic equipment in manufacturing) and automaton (moving mechanical device made in imitation of humans) systems in the sense that they are capable of making decisions themselves. The world is rapidly advancing to an era consisting of autonomous systems.
- In agriculture, farm workers have been replaced with autonomous farming machines capable of communicating with the GPS satellites, while using computer vision to detect terrain and autonomously work on fields. Another example: autonomous weed removal robots.
- In smart factories, assembly line automata robots are being replaced by cloud-connected autonomous robots that can visualize the entire production chain and make decisions on its own.
- Autonomous chatbots are replacing outsourced and in-house support. Sales mentorship is being replaced by robotic mentors who listen in on calls measuring and mentoring based on the emotionality, tonality, and sentiment analysis.
- Robotic Process Automation (RPA) with varying degrees of autonomous capabilities is replacing legacy data entry operators.
- Autonomous PizzaBots and BurgerBots in the food preparation industry are replacing pizza and burger cooks. Autonomous barista robots capable of making custom espresso drinks are replacing baristas.
- Self-driving cars, autonomous lawn mowers, pizza delivery robots, golf ball return robots, house vacuuming robots are available or imminent.
- Even traditional machine learning is becoming more and more autonomous with self-supervised learning, transfer learning, and meta-learning. AutoML (Google), AutoKeras, auto-sklearn, and EZDL (Baidu) are competing for automated machine learning.
Why will we create jobs?
David Autor, professor of economics, in an MIT Sloan and MIT CSAIL lecture, explained we always create jobs because of O-Ring Economic Development and Insatiability principles.
O-Ring Economic Development Principle: The failure of a single gasket ‘the O-ring’ caused the 1986 Challenger Shuttle catastrophe. The O-ring was the shuttle’s weakest link. Improving the reliability of the weakest link of a system improves the reliability of the entire chain. Similarly, improving the weakest link in a process consisting of humans and automated tasks improves the whole process. Making tasks automated, in turn, increases the reliability, efficiency, and overall cost. Automation thus increases the value of humans in a collaborative process consisting of humans and machines.
Insatiability Principle: With technological advances and rising wealth, we take better care of our necessities, and think of new things to engage our attention, occupy our creativity, and use our time. So when we invent new things, we need new things. As Thorstein Veblen appropriately said, “Invention is the mother of necessity.”
So as it turns out, we create jobs, to increase overall human value and because we need new things when we invent new things.
Is creating jobs enough?
Is creating jobs regardless of whether UBI or a variant materializes, enough?
History offers a valuable lesson. In the late 18th century, automation and advances in fertilization, irrigation, genetics, and technology caused Agricultural employment to shrink. In 1900, 40% of all US employment was in agriculture. In 2000, 2% of all US employment was in agriculture. Agriculture became highly productive and, a few million farmers could feed the entire US population.
With this efficiency and productivity, the farmers’ children were no longer needed on the farm. Unfortunately, they weren’t really prepared for other industries. As a result, the US mandated that children stay in school until age 16 and this mandate turned out to be the most significant public investment in US history. The “High School Movement” resulted in a country with the most skilled, educated, and productive workforce.
Besides education and job creation, we will need a Movement to prepare people for the autonomous era. We will need to help them become successful in the autonomous era.
We may need a UBI like system for reducing income equality, but we will make much more progress by creating era relevant jobs, along with education and appropriate ‘movements’ or nudges.
Would a non-government UBI variant work better?
Some interesting articles
The 2013 Future of Employment Oxford research paper ranks jobs and the probability they will be replaced by autonomous systems. Bloomberg looked up the Standard Occupational Classification codes and has converted the data to a palatable interactive chart.
MIT Sloan and MIT CSAIL Lecture by Economics Professor David Autor, – (M5U2 Casebook Video2) – awaiting permission to share video or transcript
See Professor David Autor’s article: Why Are There Still So Many Jobs? The History and Future of Workplace Automation