AI-Driven Economic Growth Is Coming: Are We Ready?

---

• Humans are innate tinkerers and technologists, says Mark Mills, who heads up a national energy think tank. Technology makes civilization — including laws, government and politics — possible.
• Even if the convergence of different technology-driven forces isn’t unique, the current convergence of information technology, materials and machines is.
• Will America sovietize its economy? Probably not, and — despite somewhat constrained growth — economists are surprised that the U.S. is surviving the predations of policymakers.

If you’re wondering where productivity boosts are going to come from in the economy, you’re not going to have to wait much longer.

You’re looking at an imminent and brand new Roaring Twenties, says Mark Mills, Executive Director of the National Center for Energy Analytics. Mark is the author of The Cloud Revolution: How the Convergence of New Technologies Will Unleash the Next Economic Boom and A Roaring 2020s. He believes that there are unique opportunities to grasp — so long as America doesn’t hamper itself.

Mark rejoins Kevin Coldiron for a stimulating discussion about the beginning of a roaring period of productivity and economic growth, driven by technological convergence that hasn’t been seen in society for a long time.

In fact, while the concept of technological convergence itself isn’t unique, the current convergence the world is experiencing is. Those who embrace what technology can do for civilization — as well as for markets — are set to gain big time.

These are some of the key takeaways from Mark and Kevin's discussion about where the economy may be headed and the evolution of economic trends in light of his expectations.

To make our future is human

“We’re all forecasters and we’re always forecasting,” Mark says. “It's how we plan and live our lives.” Basically, it’s human nature to think about the future — be it business, government policy or life more generally.

While technology is a special sub-category, it impacts everything. Ten thousand years ago the world couldn’t have looked more different, technologically-speaking, than it does now. And yet, “the same energy forces, materials and nature exist,” Mark explains. “What's different is our knowledge about information and the technology we've developed from that knowledge to manipulate nature — literally manipulate.” To build, we manipulate atoms and molecules. Meanwhile, technological change and progress are different things: The former doesn’t necessarily result in the latter.

To be human is to make: “We’re innate tinkerers and technologists,” Mark says. “Technology is not exogenous to humans: It’s human nature.” Politics, governance and law all matter in making predictions about the future, but it’s “technology that informs and enables everything that makes civilization possible.”

The unique convergence of three forces

Because of the inertia of systems, broad decades-long predictions about the future are “necessarily dictated by what we know how to build today,” Mark says. But what people don’t often realize is that many revolutionary technological changes have histories stemming back decades into the past: Think automobiles, airplanes and computer chips.

“Every entrepreneur will tell you that the overnight success of 20 years of incremental engineering and proven development leads to that inflection point where that new product, technology or service emerges suddenly on the market and in the public conscience,” Mark explains. The apparent leaps and bounds in AI that the world has recently seen are a case in point.

The three forces that lead to a convergence resulting in economic growth are:

1. Information technology including microprocessors
2. Materials such as those from mining
3. Machines that we use to move and make things

The world hasn’t seen such an era-defining convergence in a long time. In the past, when world-changing inventions like the printing press and television converged with radical developments in machinery, there was a bigger revolution than if one of those things had happened individually.

“It happened in the 1920s with air travel, the car, the telephone, and the telegraph, and of course the advent of pharmaceuticals and steel — three revolutions that typify the change and happen simultaneously in different domains,” Mark says. “We have the same thing happening now: The cloud itself as an information tool is the biggest infrastructure ever built by humans,” Mark says, “and combines with AI, which is now maturing to a point of high utility.” Advances in both biological and physical material sciences — along with the development of autonomous robots and machines used to make microprocessors and semiconductors — also converge with these trends.

No surprise that tech results in rising energy consumption

On the subject of developing trends since his first discussion, Mark says there have been “no surprises” — due to the nature of inertia and the velocity of building things, which are well-established laws in physics and engineering. “You get discontinuities over long periods of time, but in a decade they’re very, very rare,” he explains.

When something like AI enters society's consciousness, stock markets react. Witness the Mag 7 stock explosions and the media taking notice of computing power demands. As early as the 19th century, the economist William Stanley Jevons knew that efficiency and production cost reduction increased consumption: Just look at steam, electricity, and compute. “The market response to consume more valuable goods is almost always greater than efficiency gains,” Mark explains.

But AI is painted as a bit of an environmental boogieman. It’s the data centers — essentially warehouse-sized computers, Mark says — that are hoovering up electricity. “Building steel mill-class facilities at 10 times the rate that you build steel mills” will do that, Mark says. AI isn’t blameless though. It accelerates the process in two ways:

1. AI chips use much more electricity than conventional chips, in the same way starting and permanently running something like a power plant rather than episodically like a car will use more energy
2. AI increases the value of conventional data because collecting more data fuels AI

“The global cloud uses twice as much electricity as the country of Japan does for all purposes,” Mark highlights, “and more energy than global aviation — while cloud AI is accelerating faster than global aviation.”

That’s not to say AI chip-makers like Nvidia, AMD, and Intel haven’t already begun improving energy efficiency. But this will lead to more energy consumption — not less.

Nonetheless, constraining AI — via public policy, in a bid to reduce consumption — means constraining the whole economy. Having enough electricity is only a constraint if markets are constrained in supplying electricity. “This is not a size forecast — it's a behavioral forecast,” he explains. It can easily be rectified by building what we know how to build “to fuel the economic revolution that AI offers”

Framing the materials and mines that fuel cloud growth

For materials, Mark takes a realistic perspective. “Numerical framing is important,” he says. “My forecasts are not based on what I'd like to have.”

Mining is at the front and center of his theory: “There is nothing in society — no product or service that exists — that didn't start with a mine somewhere.” The exception is in primitive societies, wherein animal-based materials like bone, sinew, and leather prevailed. And even with some recycling, mining virgin material is impossible to avoid. “Everything costs energy; everything involves waste,” he says.

Starting with mining, the questions become about quantities required and cost of material.  “Making the complex things that we build and operate takes energy,” he says. When they become waste, they're worn out, broken and fall apart — the energy went into their complexity, not just the materials.” It’s Physics 101, Mark says.

Physical automation is also set to improve mining through “robotics, better knowledge of how we can manage materials and more efficient extraction” via enhanced chemical processes. AI-driven supercomputers will improve these processes further. “All that will happen, but it takes time — these are industrial systems,” he says. “They get better measured in decades, not years.”

Finally, framing mining has to start with understanding reality, not wishful thinking. “Mining is not an easy industry,” he says. “The average time to open a new mine is 10 to 20 years,” Mark says. In the U.S. it’s longer due to opposition, so few if any mines have opened for decades, despite there still being mineral resources to mine. “We've chased mining out of the United States,” he says. “It didn't leave voluntarily.”

But there'll have to be more mines for economic growth “Acknowledge that growth is good and requires more materials,” he says. “You want to extract them in the most benign way possible.” Better here than in other countries with more lax labor and environmental laws and protections.

“Once you’ve framed it right, you might arrive at policies that are sensible,” Mark says.

Manufacturing as a service

What brings together the advancement of the three different dimensions is the notion of manufacturing as a service (MaaS). Mark offers two different examples to explain:

1. Nvidia has no manufacturing plants for its AI chips — the MaaS is provided by Taiwan Semiconductor Manufacturing Corporation (TSMC)
2. The prevalence of specialization and internal integration varies between industries, and while its evolution is unavoidable, AI, automation, and cloud technologies accelerate the process

Mark predicts an increasing number of non-semiconductor companies adopting an Nvidia model “where they design, vet, and market things that they don’t actually manufacture.”

Most people have never heard of publicly traded Canadian Magna International, but it manufactures world-class parts for BMW. Likewise, “most vintners don’t grow grapes,” Mark says — they buy them. It’s now easier — and more important — to specialize in manufacturing due to automation and AI.

The economy is resilient

With interest rates up from zero to 5% in 18 months and no economic crash in sight, Mark believes that “we're in a period similar to the 1990s: We can see computers everywhere except in the productivity statistics.” While hard to measure their ultimate effect, it bore out soon enough.Likewise, next-generation technology computing is currently hard to tease out but is already making waves.

“There are data supporting the contention that a lot of the resilience in the economy is in fact coming from AI and cloud technology,” Mark says. This is the beginning.

“The extent to which we can really roar and get big productivity boosts,” he highlights, “is now going to be significantly impacted by policy.” Constrain or unleash: This is the choice facing policymakers. But Mark believes it will happen anyway because markets are unstoppable in the long run.

“I'm hoping and don't believe America will sovietize its economy,” he says. “We have somewhat constrained the growth, but not as much as the partisans on both sides would say or complain. Hence, the economists’ sort of surprise that we're surviving the predations of policymakers.”

Here’s to hoping we continue to survive them.

---

This is based on an episode of Top Traders Unplugged, a bi-weekly podcast with the most interesting and experienced investors, economists, traders and thought leaders in the world. Sign up to our Newsletter or Subscribe on your preferred podcast platform so that you don’t miss out on future episodes.