AI and the Limits of Adaptability

How will AI roll out and what will it mean for humanity and for your business?

• This year, enterprising executives seek to automate tasks like data entry and customer support with AI tools like Chat GPT.
• Next year, Stephen King will use AI to write a book in the voice of Ernest Hemingway.
• In 2029, a snarky college student teaches a room full of monkeys to use AI to write Shakespeare.
• In the 2030’s, robotics become critical to home healthcare.
• By the 2040’s a population inversion across the 20 largest economy forces robots to become larger producers of economic value than people.
• By the 2050’s the two and a half century trajectory of the extraction economy is in crisis. Humanity must learn to deal with the limits of resource extraction or face an existential crisis. AI is the central technology that makes the necessary innovation possible.

None of these projections should be surprises. They are based on established trends in technology, demographics, and globalization that have been studied and written about for years.

We know that AI is improving; we know the world is getting older and more urbanized; we know that trade is globalizing at breakneck pace. If your job includes any sort of planning for the future, you’re probably intimately familiar with all of these trends (and if not, you probably ought to, quickly).

But these trends are moving into uncharted territory, and pushing the systems that power our civilization to the edge of their adaptability. 

Until the advent of air travel, global trade was limited by the speed of ships. Until the microprocessor came along, computing power was limited by human capabilities. Every system is limited by the technologies it depends on, and cannot expand indefinitely unless the technology changes. In the next several posts, We'll explore four systems that are rapidly approaching their limits.

1. Economic and Political
2. Technological
3. Productivity
4. Demographic and Skilled Labor

Economic and Political Limits

To begin with, the system of international trade and political power that’s existed in something like its current form since the early 20th century, is reaching an end. In this system, “global trade” meant trade between a handful of economies, mostly in Europe and North America, joined more recently by Japan and industrializing Asia. It also assumed a sharp divide between the First and Third World, the former possessing nearly all the capital and political power, the latter being a source of resources and a playing board for political influence. 

This is all going away. The Brookings Institute projects that 2.1 billion people will enter the global middle and upper class by 2030, 87% of them in Asia. That means wealth, production, and power are decentralizing, not just in a rising China, but in growing markets and production centers all over the planet. Trade between a few countries is becoming trade among dozens of countries, and this is something our current regulations and financial systems are still only vaguely aware of. 

Part of this is just the inevitable fruit of global development, as poor countries become less poor and start catching up to wealthier economies. But much of it also comes from a global shift in business, away from goods (which are expensive to move across borders) and services (a bit easier to move), to data (which moves freely). If you want to buy some data, whether it’s an article, a movie, a marketing list, or a virus script, it hardly matters where on earth it’s coming from. This means that any country can participate in the trade, and that nation-level regulations on such trade are nearly meaningless. The sooner we recognize the brokenness of the old system, the better.

Technology Limits

Second, we’re facing a series of technological limitations which will come to a head in the next five to ten years.

The first is a limit in bandwidth, which may come as a surprise to those of us still geeking out about fiber broadband or 5G. Yes, it’s true that the pipes are getting fatter. But they’re not getting fat enough, fast enough. 

The “big data” revolution that’s captured the attention of the business world over the past several years is based on a dramatic expansion in data gathering, largely because of cheap sensors and growing interest in digital data about all kinds of behaviors. Google, Amazon, and Facebook demonstrated how lucrative it can be to know a lot about what consumers and users are doing, and this sparked a rush to capture as much data as possible. Such data capture is easy to automate, and therefore relatively inexpensive, resulting in a tenfold increase in global data volume in less than a decade.

Increasing bandwidth, on the other hand, is limited by hard, physical constraints. We can’t simply decide to send more data through a pipe—we have to rip out the old pipe and put in a bigger one. As a result, there’s a growing mismatch between how much information the world is producing, and how much it’s able to move from one place to another. 

Physical constraints are also behind a looming limitation on computing power. Moore’s Law has famously given us reliable increases in processing speed and memory, as chipmakers are able to pack circuits more and more tightly onto a chip. The current size of switches, though, is getting so small that it’s threatening to bump up against the size of atoms themselves, and that’s a very hard physical limit. Quantum computing may offer a way around this, but it’s so far just a theoretical exercise. At the moment, it looks very likely that the rate at which processing power increases is due to decline, or even flatten out.

We’re also approaching a limitation in energy production, as demand outstrips supply. The world’s population is still growing, and more importantly, it’s becoming more affluent, which translates directly into increased energy demand. 10% of the world’s population is expected to join the middle class by 2035. This leaves two options: either energy production will need to more than double in that time, or businesses and consumers alike will have to define new ways of living at a modern standard, on a sharply reduced energy diet.  

Production Limits

Third, we’re facing a limit in productivity growth. The rapidly expanding global economy of the last 250 years has been driven by a combination of population growth and relentless innovation, first in agriculture, energy, and manufacturing, then in communications and information technology. Innovation was well rewarded: if you had a genuinely useful idea, you could patent it, build a company around it, and grow wealthy for years or decades before anyone else caught up. In the process you might build tremendous wealth for your employees and investors, and for the nation to which you paid taxes. 

Today, that “years or decades” has shrunk to months. The innovation window—the time between when a new idea becomes viable and when it gets knocked off by a cheaper, faster competitor—is closing. This is inevitable in a globalized economy with an abundance of potential competitors, able to see what each other are doing in real time and address each other’s customers directly. It spreads productivity out far wider, so that any individual has the potential to be more productive than ever. But at the scale of an organization or nation, it reduces the time over which rents can be extracted from innovation, causing the rate of growth to taper off. 

The only exceptions, as we’ll see later in this book, are the organizations and countries that grasp these new realities first.

Demographic Limits

The last and possibly the most consequential limitation is the availability of highly skilled labor. Innovation is now table stakes for any growing company, and innovation depends heavily on people with the education, skills, and insight to create new tools, processes, and products. 

In developed economies, about half of all economic growth is the result of people and half the result of machines. In emerging economies, about 2/3 of productivity growth is the result of new people entering the workforce–a fraction that will shrink as they develop and rely more on automation. But as accelerating innovation shrinks the time that companies can use it to differentiate, the importance of labor will continue to rise, particularly highly skilled labor.

At the same time, birth rates are dropping in nearly every country on earth, creating a massive skilled-labor shortage. This creates a massive and unavoidable threat to economies that have spent the last couple of centuries enjoying sustained growth through a combination of increasing innovation, education, and population. 

In a society that runs on well-trained engineers, programmers, analysts, managers, and policy-makers, a shrinking pool of educated labor means increasing wages, often at unsustainable rates. By 2030, it’s estimated that there will be a global shortage of 85 million high-skilled workers, pushing global annual wages up beyond inflation by an average of $9500 USD. For organizations that rely on high-skilled labor, this adds roughly 20% to their number one business expense.

Taken together, these limitations pose serious threats to businesses and government entities, not because they reduce productivity, but because many organizations will fail to adapt to them rapidly enough. Everywhere we look, our ability to adapt to change is lagging behind the speed at which things are changing. Clearly, something needs to change dramatically.

Please comment, share and subscribe!

I write this newsletter to meet cool new people like you, and discover what you think. Please share and comment! Oh, you’re not a subscriber yet? Scroll back top of the page and bink that blue SUBSCRIBE badge. You will receive a notification when each new edition drops. It couldn’t be easier.