One Robot Per Child
You think things are changing fast today? Just wait till AI gets a body in the real world.
Imagine for a moment two colleagues, Casey and Jordan. They both have jobs that require more or less the same physical and mental skills. Casey is stronger than Jordan, but Jordan is usually more jovial and inventive. Jordan works eight hours a day, five days a week, and occasionally does some overtime. Casey doesn’t have a family, while Jordan is dating someone and hopes to start a family one day.
Who would you hire? Obviously you don’t have enough information to make that call yet since it will depend on the job description, references, salary expectations and many other factors. But what if I told you that Casey literally never takes time off. They work 24 hours a day, 7 days a week. What if I told you that they had the same level of physical dexterity and ability as Jordan, but that they had at least twice Jordan’s physical strength. They learn really fast. You just tell them what to do, give them a manual or demonstrate a complex task and they remember everything perfectly. Once trained they do what they’re told, over and over. And they never get tired. As in never, ever. Never complains either. What if I told you Casey was an AI powered robot.
While Casey the robot does not exist yet as described above, humor me and let’s just pretend it does. I spent more than seven years leading an AI and robotics project at Google and have become convinced that Casey the robot is coming. Not tomorrow, and not the next day, but it won’t be that long until increasingly capable robots that work fully autonomously alongside us begin to show up in our workplaces, and eventually in our homes.
Robots already exist in many different forms and I believe we will see an explosion of designs in the future. Some people believe that human-like or human inspired versions, often referred to as humanoid robots, are the best designs. For instance, Tesla is working on a humanoid robot called Optimus, Boston Dynamics has demonstrated robots doing programmed jumps and dances for a while now, and there are many start-up companies such as Agility Robotics, Figure AI and 1X trying to make robots do interesting things. I don’t actually think it’s that important what the robot looks like. Feel free to imagine your own idea of what Casey the robot might be like as it works alongside you. Does it have legs or wheels? Does it have one arm or three arms? Can you speak to it? Can it dance? What would make it feel safe and be capable?
Once we take the leap and imagine your idea of Casey working alongside us, what happens next? Casey is polite, a fast learner and an incredibly hard worker? When one Casey learns something, all the Casey’s can automatically learn the same thing? Casey will courteously and effortlessly interact with people, take commands and be perfectly capable of doing almost every physical task that people can do? …and then some. Perhaps Casey would not be the first companion that would come to mind when Jordan heads out with their co-workers to grab a few beers after work — in fact it would still be back at work, working — but if it was able to perform all or most of Jordan’s tasks, autonomously, what would that mean for Jordan’s job and labor as we know it?
AI and robots are transforming our relationship to intellectual and physical work.
Since the introductions of OpenAI’s ChatGPT, Google’s Gemini, Anthropic’s Claude, Facebook’s Lambda and a plethora of other generative AI systems, much attention has understandably been focused on the impact on jobs or tasks that traditionally were not seen as being directly threatened by technology and automation. We are suddenly asking what will the effect be on creative professions? On storytelling and art? On teaching and learning? On journalism? On science and engineering? On programming? On the law? On finance? On medicine?
Ironically, in the shadow of the recent AI breakthroughs, the jobs that require physical skills and dexterity appear somewhat “safe”. And while this may continue to be true in the short term, soon the physical world will have its ChatGPT moment, with some incarnation of Casey-the-robot, powered by AI capturing our collective attention and radically changing the nature of physical work at large. Having recently app-hailed my first Waymo self-driving robo-taxi ride in San Francisco — it was a great customer experience — I am more convinced than ever that the technology powering robots is getting ready for prime time. Most types of labor will invariably be affected.
Is work just evolving or is it completely different this time?
There is a thriving debate about whether artificial intelligence and robots are going to have a different impact on the economy than the mechanization and automation of work has had over the past 250 years since the start of the first industrial revolution. There are two schools of thought. The first is based on the theory that work doesn’t go away, it just changes. The second argues that this time it’s different because we are automating intelligence and the machines are therefore poised to take over an ever expanding set of uniquely human tasks, keeping us around as pets if we’re lucky.
The invention of the steam engine in the mid 1700s amplified human muscle power. A dizzying and accelerating rate of technological inventions and innovation ensued all the way up to the modern-day computer drove automation and efficiency to levels that were impossible to imagine at the beginning. This led to rapid economic growth, a steady increase in global prosperity and continuously evolving forms of employment across the globe. We have witnessed an amazing increase in standards of living and average human life expectancy almost everywhere.
At the same time the human population exploded, growing from one billion at the start of the industrial revolution to more than eight billion today. It would be impossible to sustain eight billion people on this planet without all the technological inventions and innovation we’ve seen the last two centuries. Consider sectors like farming and food production, transportation, energy production, and manufacturing, all orders of magnitude more productive today than they were two hundred years ago. What is considered work today are mostly tasks that people would not have been able to imagine a few hundred years ago. It is natural to assume that this evolution will continue. People have always “worked”, but not always the same way. New technological innovation and invention will unleash new forms of work and give people new ways to make an income in the future just like it has in the past.
Two factors make it reasonable to assume that things will be fundamentally different this time. The first is that certain technologies, AI and robotics core among them, may be on the cusp of radically changing the relationship between human and machine? The second is that the social and environmental cost of how we are operating is too high to sustain us going forward.
What got us here, won’t get us there?
There can be little argument that technology has been the driver of enormous economic growth and rise in prosperity. It has made us enormously more efficient the last two centuries, yet the social cost has been incredibly high as well. The incredible innovations that have driven these productivity increases have also created huge dislocations and we have become aware of the large negative impact on the environment and climate. More and more people are asking whether we have reached a turning point where the economic system that created the foundation for the value creation we have seen the last two centuries may have run its course? Can capitalism, as it is currently practiced, steer us through this next phase? From Bill Gates, to the consulting firm McKinsey & Co, to the Brookings Institution, and countless others, people are questioning whether capitalism as we know it today needs a major rethink and overhaul.
Our current economic system is predicated on a model where the social cost of growth — the long term costs to society — is more often than not not factored into the equation. What is happening with the environment and with the climate are examples of where we are beginning to witness the consequences of the increase in prosperity a majority of the world’s population enjoy today and the enormous costs future generations will be paying for our past choices. In The Value of Nothing, Raj Patel, an economist, stipulates that if we included the hidden social / ecological costs of a McDonald’s hamburger, it would cost $200.
If we factor in the down-stream costs of our current ways of working, it is becoming increasingly evident that business as usual is likely at odds with a future of prosperity and safety for most people. The entire value-creation and economic system that has supported the huge growth in knowledge, energy, medicine, manufacturing, distribution and consumption may no longer be viable.
Enter AI powered robots. Technology is continuing to drive ever faster and disruptive change. Casey, Jordan’s robot co-worker, may not be real today, but if it does become real in the next decade or two, and I think it will, then what? If Casey is highly efficient and cost effective, and maybe even super intelligent what happens to Jordan’s job? What becomes the nature and goals of human work? The overall rise in prosperity and increased life expectancy brought on by the Industrial Revolution has led to a dramatic change in what it means to be human. Yet we may just have seen the proverbial tip of the iceberg. And while this may be anxiety provoking, even threatening to some, I think it also creates incredible new opportunities to re-think our social and economic systems and ultimately our roles as humans.
AI and robotics fundamentally changes what it means to be human
In Why we should think about the threat of artificial intelligence the technologist Gary Marcus writes “There might be a few jobs left for entertainers, writers, and other creative types, but computers will eventually be able to program themselves, absorb vast quantities of new information, and reason in ways that we carbon-based units can only dimly imagine. And they will be able to do it every second of every day, without sleep or coffee breaks.” This was published in 2013. Ten years later even the creative type jobs don’t seem safe from recent breakthroughs and developments in AI. At least not as they are currently defined.
What if Casey is not a way to complement and amplify Jordan’s abilities, what if Casey fully replaces Jordan? Casey operates autonomously. It can do many if not most of the things Jordan does much faster, more reliably and more efficiently. It has all the dexterity of a person and then some, and works seamlessly in environments occupied by people. It knows how to ask for help if it can’t figure something out and once Jordan teaches Casey a new skill, all the other Caseys can learn it too.
Casey is not evil. It doesn’t want to take Jordan’s job. In fact it doesn’t want anything. Casey likely has no awareness of self, at least not in the human sense of awareness. Likely no emotion and no desires either. It does what it’s told within its domain. It has 100% memory recall, never forgetting a face or a name. It does not eat, in fact it runs on solar energy and does not need to commute to work. It may look like a person, or not. What it looks like only matters if it makes people more at ease and enables it to do its job better.
I believe the world will see the first generation Caseys emerge in the next five to ten years. If you’re willing to give me the benefit of the doubt for a bit longer, you probably agree that this raises some fundamental questions. And opportunities! Why should Jordan work if Casey can do their job better? What will Jordan do if they are not working next to Casey? Is Jordan going to become unemployed because of Casey? Who realizes the value from Casey’s work? Does Jordan receive any of that value? What happens to the taxes Jordan pays if Casey does all the work? Can Jordan be happy and find meaning doing something other than “work” in the way we think of it today?
Stepping back. Thinking out of the box.
We need to step back and take a new look at how we organize for success. A radical shift is likely necessary in order to constructively and peacefully engage the nearly eight billion people on this planet. Moreover, while we have experienced exponential 8x population growth the last 200 years, all predictions show that the human population will soon start shrinking at the same time as people are also living longer than ever. This leads to a host of new challenges such as aging populations and fewer people entering the workforce. Leveraging new technologies will be the only way to navigate the path ahead.
It is natural that we look at the emergence of new technologies such as AI and robotics against the backdrop of how we are organized and working today. Yet if we consider that the way we are currently operating is at best harmful to future generations, and at worst perilous to our very survival as a species, perhaps we would be incentivized to look at things differently. What if, instead of thinking about the future of work in terms of whether machines will replace work and create vast unemployment and social unrest, we reframe the problem? What if the very goal of the new technologies was to replace jobs as we define them today? What if we changed our ideas of work, value creation and prosperity? May it be that the technological changes on the horizon are arriving just in time?
Some people live to work. Others choose to work to live. And as our lives evolve, priorities often change too. What if robots and AI were our ticket to giving people the opportunity to live more and perform labor (a lot) less. In fact, what if the idea of “work” as productive labor became outdated as something the machines do. Humans live. Machines work.
It may be that machines will do the work that makes life possible and that human beings will do all the other things that make life pleasant and worthwhile. — Isaac Asimov
It is not difficult for me to imagine a world where healthy, nutritious food is grown by machines and most of our buildings are built, printed perhaps, by machines. It is even less difficult to imagine a world where all cars drive themselves. Or what about a world where medical diagnostics are mostly done by really smart machines that can also remotely monitor your vitals and give you a heads-up two hours, or how about two years, before you’re going to have a heart attack, not twenty minutes after.
Our work need not define us as humans. What if we could focus on what it means to live a meaningful life? What does it mean to make meaningful contributions to your community? What does it mean to be a good citizen, a good friend, a good parent? What does it mean to be human?
Today’s conventional wisdom is that free markets, competition and growth forms the foundation of our incredibly successful system called capitalism. It has proven to be a force for good and a driver of prosperity. It has become a common refrain that a model where machines not people do most of the work would be demeaning and degrading. People would lose their sense of identity, dignity and purpose if they do not have regular jobs. What would we do, hand people a check every month so they could survive while the machines do the work? Do we really want people to live on hand-outs? It is a recipe for disaster and will lead to apathy and even chaos. So goes the conventional wisdom. I posit that this wisdom may turn to folly.
One robot per child
Let me end with an out-of-the-box thought experiment and conversation starter for how robots and humans might co-evolve in a way that could create enormous economic value, be a lot more sustainable than how we live today all while freeing people up to live fulfilled lives.
What if everybody were awarded a proxy robot? Every adult person would own a robot. Think of it as your very hard working, highly capable real-world robot twin. Your robot would exist in the physical world and it would also have a virtual twin; like a personal supercharged ChatGPT agent that is a representation of you.
Perhaps you would get your proxy robot when you turn 18. You can use your robot to do whatever you want. You can send it to work for you. You can keep it at home, cleaning the house and watching your children while you are working, doing community service or training for a marathon. You can teach the robot to have unique abilities. You teach it to become your proxy, your agent in the virtual and physical world. Occasionally it may ping you and say “I’m stuck, how do I handle this situation” or “solve this problem”? Your responsibility could be to help it out, teach it more skills. Pretty soon the robots will know more than we do and become our teachers, amplifying learning and knowledge in radical new ways.
You would get your robot for free. It gets free maintenance, free power — remember, it’s charged by the sun — and free upgrades. Even better, whatever money your robot makes, is yours. You’ll pay taxes on your income just like today, whether it’s brought home by the robot, by you, or both. It would likely become necessary to regulate robot ownership? For example, robots would be non transferable. A vibrant robot service marketplace and economy could evolve. Robots and their virtual agents will work in factories, hotels, hospitals, schools, call centers, retail stores, restaurants, offices, banks, creative agencies… pretty much anywhere people work now… and then some.
Dinner is ready
Casey just got home from work and is telling me that my delicious, nutritious, healthy zero footprint dinner is ready so I’ve got to wrap this up.
Let us think out of the box. We need to turn things on their head. AI-powered robots and virtual agents are not inherently bad, but they will be inherently disruptive. Now is the time to prepare for a future where what defines us as being human is inevitably going to change.
I wrote the first draft of this post in early 2016 after I joined Google X to lead Everyday Robots, a secret robotics project, a marriage between real-world physical robots and artificial intelligence. The post was an early attempt at trying to make sense of what might happen when robots and AI one day become as capable as people.
A few questions that motivate me are; How can we free people to do the stuff they are uniquely good at and create (artificially) intelligent machines to do the rest? How can we relieve people from harmful and tedious, repetitive work? Why shouldn’t we be able to be students of any subject we are interested in for as long as we want? Why isn’t being a friend and building social relationships in the world around you as valuable as the kind of job you have today? Why isn’t being an artist who challenges norms, asks fundamental questions and creates beauty and mystery as coveted and financially rewarding as managing a hedge fund or being a bond trader? Why isn’t dedicating your time to community, family and child rearing as valuable to society as working in the public or private sector? And, how might we enable all this to increase sustainability and decrease our footprints, all without impacting societies and the people who live in them in harmful ways.
Please let me know what you think?
