The Giving AI: Abundance & Post-Scarcity
Thomas Larsen
You might be reading this from a phone in your hand, no wider than a brick and thinner than most books, and yet it has the ability to unlock the sum of the world’s knowledge to you. Have you ever stopped to think about what made something like a smartphone even possible? You can think about the materials in the phone, the metals and glass and silicon. Maybe you’d think of the technologies represented in the phone, the different sensors, cameras, microphone, or computer chips, the LED display, the ports and buttons and so on. You may have even thought of Moore’s law, the doubling of information on a single chip every few years, but you’d have to look even further back than the beginning of Moore’s law to begin to see the true start of the progress that led us to the technologies we take for granted today. The Law of Accelerating Returns, a term coined by Ray Kurzweil and often compared to Moore’s law, describes how innovative technologies today build on the sum of all technological innovation that precedes it (that’s 3 million years of technological innovation!), and how the rate of technological innovation increases exponentially over time. This exponential growth of technology starts off slow, innovation being practically imperceivable at first, but, according to the law of accelerating returns, any innovation serves to further fuel subsequent innovation. This law was recognized by Newton when he said that any thing he did for the progress of science was due to the fact that he was “standing on the shoulders of giants”. While Newton was talking about scientific discoveries, the same goes for technological progress (the two almost always go hand in hand). Many of the past paradigm shifts in technology are buried in history, but they shaped the modern world in countless ways.
The overarching goal of technological progress has always been to improve the survival of humanity by improving our collective access to vital resources. This goal continues today, whether we are aware of it or not, even though the resources that we deem vital look very different to the vital resources of our distant ancestors. The end goal of technology has always been, when drawn out to its logical conclusion, equal access to unlimited leisure and unlimited resources. Or, in other words, building new tools to do all the work for us so we don’t have to. This end goal has been unattainable for nearly the last 3 million years of trying, since our first ancestors began using broken river rocks to cut, stab, and smash, but we may be finally beginning to approach the realization of this goal.
To see how the wheel of meaningful progress has sped up over time, let’s take a quick walk through some of the most important tools our ancestors created to get us to where we are today.
2.6 million years ago, the first stone tools were produced by our ancestors homo habilis, unlocking the use of previously unavailable resources, marking the beginning of the slow exponential growth of human technological progress. Over a million years later, another member of the genus “homo”, homo erectus, began to control fire, granting access to previously inaccessible calories and directly contributing to our ability to travel outside of our natural habitat for the first time. Around 170,000 years ago our ancestors began making clothes, unlocking our ability to survive in colder climates, once again leading to the second and lasting migration out of Africa. 12,000 years ago, as our ancestors spread around the world, we stumbled upon agriculture and began developing permanent settlements on every habitable continent. 5,500 years ago a mesopotamian potter discovered the wheel, improving his ability to make fine pots; 300 years later the chariot was born. Around the same time in the same part of the world, the invention of writing made the storage of information outside of a human mind possible, revolutionizing how we store and inherit information. During this same time, the art of metallurgy, or the combination of two or more metals through processes of heating, was born, leading to the development of a large array of tools, paving the way for vast innovation and invention. 4,000 years ago, people began attaching wooden instruments to their cattle to till their land, inventing the first plows, leading to greater food productivity. Around this time, the sailboat was first invented, leading to quicker travel and greater inter-community cooperation and communication. 1,000 years later, the Minoans built the first aqueducts, perhaps mankind's first meaningful attempt at geoengineering. Skip forward through the dark ages until 1436 when the printing press was invented, revolutionizing mankind’s access to information. Then, a little over 150 years later, in 1608 the telescope was invented, giving us our first glimpse into the heavens. A hundred years later, the steam engine refined travel and the function of machines. Another hundred years and we have the telegraph, sending information through wires over long distances. Then, fifty years after that we have the lightbulb, changing our relationship with time. Five years later, the automobile, which still defines humanity’s favorite mode of transportation. Fifteen years later, the first airplane flies. No less than 45 years later the transistor lays the foundations for all modern electronics, including computers. A little over 20 years later and the first personal computer is released. Less than 15 years after that we got the first cell phones, and a few years later, the World Wide Web. Now, only sixteen years ago, we got the first smartphone, combining telegraphs, transistors, the printing press, the internet, and the knowledge behind all other inventions into an object that fits in the palm of your hand. 11 years ago, we got CRISPR, the technology that enables us to easily alter genetic code. And, today, we have the ongoing effort to build helpful Artificial Intelligence (AI), which, if we are lucky, will be the way we finally succeed at our goal of abolishing the need for human labor to supply the necessary resources for human survival.
This last point may seem far-fetched, but let’s take a moment to look at a few of the ways AI could bring about this highly coveted society of abundance, which I would call a limited post-scarcity economy; “post-scarcity” simply implying abundance and “limited” because it only supplies the sort of abundance of resources necessary for human survival, where a true post-scarcity society would mean abundance of all resources (which is arguably an impossibility).
AI is very good at doing repetitive, well-defined tasks. The more repetitious a task the better suited it is for AI. This holds true across all types of tasks, markets, disciplines, and business models. There is no repetitive task no matter how difficult that cannot currently or very soon be done more quickly and efficiently by an AI than by a human. This shift from human repetitive behavior toward robotic and AI-driven tasks is seen across the board from manufacturing to farming, from marketing to digital art. For years now we’ve used algorithms to adjust and fine-tune marketing campaigns, GPS-driven combines to harvest and process agricultural products, robots in car manufacturing, and now programs like DALL-E and Midjourney to create digital images. I have personally found ways to offload some of my more monotonous and repetitive tasks to AI systems. Rather than search the internet for someone who has had a parallel problem to solve in excel and written out their solution clearly enough for me to follow, I can simply ask chatGPT and it instantly gives me step-by-step instructions on how to solve it. Soon, both Google’s and Microsoft’s new AI assistants will be able to simply act on my request and do my spreadsheet work for me.
On the one hand, this technology (AI) has been developing for only half a century (starting with the summer workshop of 1956 with John McCarthy, Marvin Minsky and their friends), but on the other hand, it is simply the next step in our 3 million year trend of exponential technological growth. Soon, however, this exponential trend will go through the drastic change expected of exponentials; our innovations will begin to do the innovating for us, sparking that dramatic near-vertical uptick. Coding of new systems has already been greatly augmented by current AI systems in the past few months, and soon we will simply give these systems a task, and it will code a solution to that task. Then, after that amazing breakthrough, we will see a time of self-prompting and self-coding based on general initiatives, like “solve the climate change issue” or “develop a warp drive.” There is already a version of GPT4 called AutoGPT which is capable of self-prompting and working towards multifaceted solutions to real world problems. This groundbreaking proof of concept popped up mere days after GPT4 became available to developers to tinker with. Innovation rapidly builds on innovation.
All this is to say that the automation of most jobs is nearly inevitable, and at that juncture we have two main ways in which we can react:
1. We leave things as they are for as long as we can, we all lose our jobs, the rich become exorbitantly more powerful and the world becomes unlivable for most, or
2. We find a way to redefine some needs as rights that ought to be guaranteed by society, and employ our new technologies to achieve that goal.
Of these two options I think the second option is preferable, and I imagine most would have that same preference.
Let me briefly describe to you what could be if we allow ourselves to replace greed with human optimization. Our current economic system of government-regulated capitalism is neither a free market, nor optimized for human well-being. An AI-assisted post-scarcity market could not only help us set up a universal basic income like past-presidential candidate Andrew Yang and several major players in the technology industry have insisted on, but it would set up a universal basic right to healthcare, shelter, food, education, and income. It may even be arguable that we already have the means to achieve this level of universal rights currently, but in a foreseeable future where AI and automation remove the need for human labor to supply us with the means to secure those basic needs, a true sense of basic human equality can be achieved without being built on the backs of an unseen class of human workers.
With everyone’s basic needs taken care of, all of humanity would be free to pursue their passions. Many would choose to go to school to learn, many would choose to teach and mentor, impacting generations of learners. Many would pursue art, the sciences, engineering, or become nurses and doctors. Most, if not all, would choose whatever profession they most want to pursue simply because it makes them happy, adds value to their lives, and serves to better the conditions of human life in general. Today, many people quit school or never pursue it to begin with because they need to work to support their families, some people pursue careers in fields they hate simply because of the paycheck, and nearly everyone sacrifices some level of mental, emotional, or spiritual well-being in order to simply survive. This would no longer be necessary. If we choose to guarantee basic needs as fundamental rights, we will open the human race to a level of satisfaction and purpose unseen in any previous generation.
I often hear at this point the common question, “if people aren’t required to work in order to survive, won’t everyone just become lazy, watch tv all day, stay home and play video games non-stop?” And I grant that this is indeed a possibility, especially at first, but if you stop and imagine what you yourself would be doing with your life if money, shelter, food, and so on was guaranteed and not something to worry about, I imagine that answer is not “sitting at home and playing video games”. In fact, I think many of those activities exist simply as outlets and distractions from the monotony of unimportant jobs we don’t actually want to do and the subconscious depression over our dissatisfaction with how we live our lives toiling endlessly to achieve the goals of others just so we can spend nights and weekends doing some fraction of what we want.
What about service jobs? Who would do those? There are companies currently working on machines and their accompanying AI systems with the dexterity and agility to achieve greater than human level object manipulation. If they succeed, any manual task we currently do could conceivably be performed by machine. But, setting aside machine capabilities, there remains within many of us the innate desire to tinker, and fix. I have a friend who is an engineer working in the aviation industry. He is well paid and generally enjoys his job. When asked what he’d be doing if money wasn’t an issue and all of his needs would be taken care of no matter what he did, he said he would spend his time fixing things for people. He loves solving problems that require a technical mindset and loves helping people. This example, though obviously anecdotal, points out the potential for a great shuffling of careers that would inevitably happen if these rights were guaranteed. Many high-paid individuals might take up professions that require manual labor like gardening, mechanical work, and carpentry, while others who would have had less opportunity in our current system might go to school and pursue higher education and careers that require years of schooling. In a society like this, people would be drawn to the professions that best fit their passions and capacity rather than their desire for more money. Personally, I would be pursuing a career in astrobiology or physics as that is what interests me the most, but I have a family, and I have bills, and I have no financial way to pursue something like this in our current system with the time given me. As the world shuffles jobs around to optimize satisfaction, we will begin to see all the remaining areas where automation ought to be utilized and AI could help us develop the new technology needed to fill in those gaps.
When discussing a system like this, I have also been asked “what happens to the free market?” In response I always ask “how free is our market today, really?” When poor business decisions were made leading up to the 2008 recession, what happened to the industries that pushed us to that failure? The government bailed them out. Why were smaller companies left to expire in the wake of that disaster while the big companies weren’t allowed to fail? But that’s not the only way the government intervenes in industry. How much do you pay for a cheeseburger? Much less than you would if beef wasn’t so heavily subsidized by the government. Economists can point out thousands of other ways our government tweaks the economy at will. We pretend in America that we believe and follow free market economics and that the so called invisible hand guides the market towards greater equality and prosperity and security for all, but in reality our market is a highly manipulated system that is being shaped for the often nefarious purposes of those governments, companies, and individuals with the pushing power to do so.
But, you might be thinking, “how would a post-scarcity economy solve this problem? Wouldn't the government have more control over the economy?” Well, it’s complicated. A truly free market holds great potential for motivating innovation and furthering the advancement of good in our society, but it needs to shift from a tool of greed and power to a tool used to achieve greater self satisfaction and to improve the wellbeing of humanity as a whole. In other words, we would be shifting from a world motivated by fear (fear of not having enough to survive) to a world motivated by love (love of self, love of family, love of society, and love for the well-being of the world).
The benefit of maintaining competition as a motivator in our future economy is that it allows people who do have a desire to live a more affluent life a means whereby they can do it. It also motivates people to continue to innovate and react to the innovations of others. But competition should not be seen as antithetical to cooperation. In a post-scarcity economy, trade secrets need not exist, patents would become unnecessary, thereby ensuring that any innovation that is developed necessarily improves the wellbeing of those you compete with and gives all a leg up to push for further innovation. If all of the currently available AI technology was completely open source, there are brilliant minds who are not currently working for these large developers who could envision ways in which this existing technology could be used to improve countless lives, but because they have limited or no access to it, that improvement may be at best postponed until someone with the means makes the same conclusions they would, or at worst might never come to fruition. Competition in the economy should be seen as friendly rather than a dueling of enemies. With no threat of the loss of livelihood, shelter, healthcare, food, etc. people would be free to take business risks that may be the leaps in innovation needed to solve some of humanity's enduring mysteries and problems.
The government's role in this sort of economy would most likely shift greatly. Where at first the government might be needed to generate the societal infrastructure that allows a limited post-scarcity society like this to be implemented, we may soon find that AI systems are much better, more efficient, more equitable regulators of the economy than people. The impartial and efficient workings of a robust AI system could lead to a decline of human activity in market governing, which could also lead to a steady decline in corruption and greed-fueled policies. In addition, we currently live in an information-rich society. We gather more data than could ever be processed by humanity. With new AI tools and the hopeful commercialization of quantum computing, the data we gather could finally be utilized to create powerful simulations of complex systems the world over, like global weather and climate and the economy, and if we grant a governing AI access to these predictive abilities, it would allow us to far more accurately predict these systems and manage our reactions to them.
In order for us to prevent the drastically negative economic, ecological, and existential effects of powerful AI, and use it to finally accomplish man’s underlying purpose for technology, some very important decisions have to be made.
First, we need to decide which of our many needs to guarantee as rights.
Second, we need to decide to use the upcoming technological breakthroughs to build collectively towards the goal of automating the work and resource gathering that physically grants those rights.
Third, we need to decide that as automation increases, human well-being is prioritized over greater profits.
With these three decisions, we hold the future state of the world in our hands. I have my opinions about which needs ought to be guaranteed as rights (as seen in the paragraphs above), but ultimately I believe these should be determined democratically. The second two decisions, however, must be made collectively. I am not a psychologist, or an economist, or any other people-oriented “-ist,” but I do know that without prioritizing people over profits, and prioritizing the development of automation technology specifically for the purpose of granting universal access to goods needed for survival, our world will remain in a state of perpetual inequality until we do. We are so close to accomplishing the goal of a limited post-scarcity society, where everyone can be lifted above the struggle for mere survival and find new meaning and new purpose in their lives. Let’s not waste this opportunity while we have it. It may not come again.