On Technology, Inequality, Purpose, and the Fermi Paradox

Walden Yan
Abstract art generated by OpenAI's DALLE 2

Everyone has seen the impact of software on productivity, as well as its ability to create a ton of wealth, albeit unevenly distributed. Software gives enormous economic leverage to individuals who know how to code. But I will explain why I think this massive economic leverage will also eventually be seen in every industry, including those that deal with the physical world, from farming to manufacturing, and why I think this will have disastrous consequences for economics and social stability. And by the end, I hope to accumulate all of this into a possible explanation as to why we appear to be alone in the universe.

Technology Makes Solving Problems Easier

As a founder, I like to build things that solve problems in this world. Recently, I've been pondering the nature of problems and the way we solve them. In my view, the core job of a business is to collect resources and target them toward solving specific problems and providing services to varying degrees of effectiveness. Of course, you can generalize this to organizations other than businesses, such as nonprofits, governments, etc. But I will focus on businesses for the purpose of this analysis. The wonderful progress of technology has enabled us to do this -- solve problems and provide services -- with increasing ease and speed. Over time, any individual job has required fewer and fewer resources. The creation of the spinning jenny enabled one worker to do the work of ten (some say even up to 120) of their peers. Modern-day SaaS tools allow one founder to do what would require just 25 years ago, an entire HR team, legal consultation, financial advisory, an executive assistant, and more.

This increased leverage of human capability, especially as we entered the digital era, as many have rightfully pointed out, has created an entirely new type of economy. For one, inequality in wealth distribution has reached completely new heights. Just in 2011, the individuals on the Forbes 400 list were collectively worth more than the bottom 50% of Americans combined. This inequality has only grown in the last 10 years. Removed from its political implications and notions of morality, this actually makes a lot of sense in the context of what technology enables. Specifically, digital technology creates economies of scale.

For thousands of years, if you were running a business and wanted to 2x your sales/production/profits, you would need to 2x the inputs to your business: employees, resources, capital, and time. To serve an entire country, you would need to build physical offices and deploy employees in every corner of the nation. And though this is still true for many industries, we are seeing this pattern start to change. In 2014, WhatsApp was acquired for $19 billion dollars by Facebook. But behind this number were only 55 employees serving 450 million users. Really take in the scale of that leverage! Each individual was personally responsible for about 8 million users. Each individual was able to create value for more people than they, their kids, or their great-grandkids for generations will ever meet in their combined lifetimes. Software scales extremely easily and at very little marginal cost per additional user served. And with the exponential growth in population in recent years, it does not seem like there will be any shortage of users any time soon.

In short, this ability to arbitrarily scale a product or service is directly responsible for a small group of people being able to have an extremely wide impact.

Is Exponential Scalability Limited to Software?

Is there something fundamentally different about digital technologies than other technologies? Kind of, but not really. Software is extremely powerful because it can be copied and distributed at nearly no cost. It is tempting to say that this sort of leverage will only ever remain within the realm of code. After all, everything else in the world still requires some sort of physical input. The nuggets and fries you get from McDonald's need real animals to be raised and slaughtered; The cars on the street need metal to be extracted out of the ground and transformed into a precise shape; The house you live in requires workers to construct it. But I would argue that not even these physical items will be safe from the grips of technological scale as humanity ages.

You might think that no matter how powerful technology gets, it cannot change the fundamental nature of the scalability of certain businesses. For instance, it seems that farmers will always need more physical inputs (land, water, nutritional supplements) to produce more food. Therefore, the upside of what a single individual can do in an industry like farming, and the ratio of business-value-to-employees must be somewhat limited. But people also forget that software businesses also require physical inputs: silicon chips, computers, data centers, etc. You may quickly forget this because now everything happens in the "cloud" (or more recently, on the "edge"). But this is all fancy terminology to make you forget that somewhere on this planet, there is a physical machine that needs to process server requests and manipulate data. There was a time when starting a tech company meant you had to buy and manage a large array of computers. The startups that did that best now run cloud computing businesses so that other people can do it without needing their own physical computers (Google, Amazon, Microsoft).

The process of acquiring computers was abstracted away once people figured out how to organize and manage these resources at scale. With this abstraction, a click of a button is all it takes to go from owning zero computers to owning 100. It is easy to organize computers in this way because they are manmade. We designed them to be easy to deal with. But when we move down the production ladder to resources farther from the touch of humanity, it becomes much harder to organize their extraction and maintenance at scale. When we look at manufacturing, a lot of the process is standardized and can be automated. But we still need some humans in the process to deal with unexpected cases and actually design the manufacturing process. Even more difficult than manufacturing is self-driving. This is a perfect example of something on the frontier of automation. No two roads are the exact same and different situations require different approaches, but a lot of it can be understood and standardized. Even more difficult is figuring out how to utilize natural resources. No two pieces of farmland are exactly suited for the same purposes. No two iron mines have the same geological structure. Humans are still very important in the processes of these industries.

Imagine the day when we can finally have a smart-enough AI -- or some other technology, but for simplicity, I will assume it is AI -- to make all of these tricky decisions for us. This AI would be able to strategize any business' resource allocation, make many copies of itself to run different parts of the organization, iterate over many different product and service ideas, and organize the logistics to make this all happen. Consider every job that really needs a person in today's world. A programmer? Not when robots learn to code. A factory manager? Not when robots learn to design and organize themselves. A farmer? Not when robots finally figure out how to run a farm on their own. Our touchpoint into all of these areas of production will be governed by these software AGIs (Artificial General Intelligence), and the entire world will become a software API. Scaling farm production could be as easy as scaling a cluster of Kubernetes is today.

What are the Implications of Extreme Scalability

Most likely, the person/people who manage to discover such an AGI that makes their business ultra-scalable will become uber-rich. It is extremely unlikely that everyone in the world will equally benefit. If it only takes one person to monopolize a business sector, there will probably only be so many well-off people as there are potential businesses to offer. If humans actually one day become multi-planetary and we see our population rise to the trillions, quadrillions, or even beyond ... is there really an equivalently exponential scale of unique business services the world needs? It seems like an individual human only requires so many things to live a good life. A single person needs on a day-to-day basis maybe 100 different services. In the future, by the time our population multiplies 10x, it is unlikely an individual human will need 1000 different services. For a large part of history, the services you needed to live were best provided by a local group of people near you. You buy stuff from your local grocery store, go to school in your local district, eat at nearby shops, etc. But someday all of the stores on Mars and all of the stores on Earth might all be run by the same AGI.

It is a curious question to consider: if all payments for services flow towards a small lucky group, what happens to supply and demand? Businesses cannot make money when their customers don't have any money themselves. So oddly enough, it isn't even clear whether those at the top will benefit. What does a system of money even mean when only a few people are actually making money? It seems that the market economic system doesn't sustain itself in such a world.

To take a very brief political detour, I think this economic instability is an important undertone when people talk about automation and social libertarian policies. The call for UBI (Univeral Basic Income) is essentially a call to take one of the first steps towards valuing work and living in the absence of the ability to provide market-competitive productivity. It isn't super clear if humanity will be able to easily make this transition. When discussing automation, people frequently point to the fact that automation of some industries leads to more jobs elsewhere. Of course, if AI is able to do any work that humans can, the only potentially irreplaceable jobs are those that require the worker to be human (for sanity, I will assume that the AGI doesn't create some sort of fake humans to do this work for it). But oftentimes these jobs require the worker to be human not because a human is fundamentally BETTER at the job, but because there is an immeasurable value of human connection. The imminent economic instability created by scalable technologies is already leading many to think about a world where markets don't effectively govern society and this is ever more important in a world where every industry is eaten up by software.

The exact solutions to these questions need much more time to be discovered. Do we need to embrace UBI? Do we go back to a communal services system? Maybe, but I'm sure there are also other solutions to be found over the generations until these problems are at our doorsteps.

Can We Learn From Aliens?

There is a sci-fi book series called the Three Body Problem. In it, the protagonist states that one of the core axioms of cosmic sociology is:

Civilization continuously grows and expands, but the total matter in the universe remains constant

This quote inspired a thought in me that potentially offers an explanation for how human life would continue to operate in a stable fashion in a world with extremely powerful AGI.

Maybe it isn't true that civilizations should always continue to grow.

After all, growth comes with obvious pains like depletion of resources, conflict, and information overload. Humans were also not made to function in the large and complex world we live in today. It is extremely depressing to try to stay on top of all the events on social media and every world conflict. It is frustrating and depriving to try to be conscious of how the other 7 billion people in the world might view you. And perhaps we have just uncovered another issue - that population growth and technological growth lead to implosive economics.

Relentless human innovation has definitely done lots of good. In the last few generations, we have lifted billions of people out of poverty. But perhaps there is some point where we don't need to keep pushing further. It is not clear to me that I would be significantly happier living in a world without scarcity than the one I live in now. Research has generally shown that more wealth, and therefore more access to material things you want, generally doesn't contribute to happiness beyond a value of about $75,000 / year. Humans like to invent, conquer, expand, and consolidate power, sometimes at the expense of our well-being and societal stability. Maybe we will eventually need to repurpose ourselves, away from this Sisyphean endeavor of "progress".

So perhaps when we look out at the stars and see almost nothing, it might not be because we are actually alone in the universe. Perhaps out there are tiny, advanced, alien civilizations who are enjoying their high-quality lifestyle with no intention to scale beyond their small population. Perhaps they have no desire to expand because it would destroy their way of life and the stability of their civilization. Perhaps it is a purely human desire to want to expand our influence to new frontiers.

A huge thanks to Robert Wachen, Rakesh Nori, and Amir Bolous for their feedback and thoughts on this piece.