AI (and companion technologies) will shape the next First Turning

Question is how, and to what end?

Previously, I wrote an article titled How central banks, gold, and cryptos will shape the New World Order. This is intended to be a companion article, — as the title suggests.

The significance of any technology lies not in its standalone awesomeness but in its societal impact in the context of prevailing economic conditions. As well, governmental encouragement of certain technologies over others is not random, but with intent to steer the economy and society in a desired direction. To understand prevailing and future economic conditions, we must first understand “how we got to where we are”.

Simply put, economic (GDP) growth = population growth + productivity growth + debt (money) growth. In modern (post World War 2) American history, the growth story can be summarized thus:

• Growth spurred by demographics (boomers reaching peak spending years in their 40s) peaked in the 80s.
• Seemingly spectacular productivity growth (PC, Internet, et al) in reality proved insufficient to make up for demographic decline.
• Meteoric debt growth (to compensate for the shortfall) had dwindling efficacy (the following is a chart of ONLY government debt. If private and enterprise debt are included, the deterioration of efficacy is worse):

A 2013 study by the World Bank found that if the debt-to-GDP ratio exceeds 77% for an extended period, it slows future economic growth. Indeed, since the GFC (Global Financial Crisis) in 2008, debt had grown (with insane money “printing”) exponentially — making for inflation in asset prices and growth in the paper financial “economy” in which those financial assets transact. But the physical, real economy — measured by wage growth — stagnated. Put another way, creation of money (debt) spurred Wall Street — but not Main Street — growth. Hence, we have the current tale of two “economies” (one real and the other financial and Bizarro) with ever-widening wealth divide between the modern patricians and plebeians. Incremental money (via creation of debt) flows directly into inflating asset prices (which only benefit the rich who own and keep assets), bypassing the physical economy (made up of the poor who earn and spend wages). In sum our current fiat (debt-based) currency system, — entirely dependent on credit expansion and currency manipulation (by the Fed controlling both price and quantity of money)—will inevitably continue to widen inequality, engender extremism, and end badly.

The crypto movement — specifically the Defi aspect of it — aims at solving the Bizarro world problem by redefining “money” (from a trust-based fiat currency that can be created out of thin air at the whims of central banks to a truth-based currency that cannot). This new currency (especially Bitcoin with predetermined upper bound of “minting”) is ideally suited for the upcoming explosion of technologies, — especially AI (discussed later) — which are by nature deflationary (because of exponential increase of productivity which dramatically drives down marginal cost of production). If Bitcoin replaces fiat currencies today, prices of goods and services will exponentially go down with exponential increase of productivity with today’s technologies. The world will no longer be captive to a system built on perpetual inflation from ever increasing debt creation (currency expansion). Put another way, Bitcoin would bring the biggest debt jubilee in history. A currency transformation this drastic poses existential threat to the traditional financial (TradFi) establishment. In response, the traditional financial establishment is now squashing the crypto movement via an engineered banking crisis. We should all pay attention to see how the plot unfolds.

Going forward, demographics (the most significant determinant to an economy and nation) will continue to deteriorate. Debt growth will inevitably resume after the current Fed tightening respite, but it will only cause further inflation and worsen wealth-divide. The only mitigation of that (and attendant social unrest) lies in productivity growth via technological innovations. And government and enterprise alike are desperately in search of “the next big thing” since the Internet.

Sure enough, the current exponential growth of Generative AI (thanks to rapid adoption of GPT4 by Microsoft and countless others) fits the bill. Marc Andreessen certainly thinks so. Generative AI has clearly emerged as the nexus for companion technologies like robotics, quantum computing, visualization technologies, Stable Diffusion and others. A more dramatic statement would be “AI is the oxygen for the next Cambrian Explosion of technologies”. Obviously, combining exponential growth in knowledge with exponential growth in computing power and data will spur phenomenal productivity growth. Robots driven by AI will no longer be limited to repetitive tasks but will become physical (not just digital) learning machines that constantly self correct and improve at the speed of light when taking on a plethora of new tasks. AI will vastly improve the efficacy of big data analytics in virtually every industry. Further, blockchain — viewed as the operating system for the Internet of Value — can be aided by AI to quickly create a truth based economy, replacing our current (fast diminishing) trust based one held captive by the finance “industry”.

Tragically, this trend will also — perhaps all too quickly — turn into existential threat for humanity:

• AI is driven by “Big Data” and lots of computing power (why Nvidia is so hot for its powerful GPU). Specifically, AI and quantum computing both solve problems probabilistically rather than deterministically. They are thus perfect compliments to each other and their combined use will surely be synergistic. But their creations will be so varied, unpredictable, and speedy that decision making will very quickly leave the human realm (i.e. AI will making decisions for humans in short order). Any talk of “regulation” at either government, enterprise, or individual level will be moot. AI (all left-brained and no right-brain, — therefore devoid of “humanity”) will bring about the much feared singularity much sooner than we can anticipate.
• AI and quantum computing, for their combined power, will quickly bring about geopolitical conflicts of a scale never before seen. “He who dominates AI dominates the world”. This will bring about a race for resources, data, technologies, ultimately leading to arms race and kinetic wars.
• Regulations and enforcement always lag perpetration of misdeed; this is the “closing the barn door after the horse had already left” or “fighting yesterday's war” syndrome. The more powerful the tool of misdeed (in this case AI technologies), the more severe and widespread the damage, and the more moot the compensatory measures. The initial use of AI to improve deep fakes, hacking, impersonation, phishing etc. are just appetizers for much worse AI-assisted exploitations to come. This clip merely purports to be an AI-assisted robot playing ping pong (it is actually CGI superimposed on a human player). Many take it at face value and draw the conclusion that AI will obviate humans in competitive sports. This type of distortion of truth, with malice of intent, can wreak havoc on a large-scale.
• The substitution of machine knowledge for human knowledge— in virtually every profession and industry — will prove too abrupt, pervasive, and rapid for either government or enterprise to ensure “alignment with human interest”. In the late 1800s, agricultural labor got obsoleted by the industrial revolution and farmers in the hinterland had to “re-invent” themselves as factory workers in cities. In the late 1900s, industrial workers got obsoleted by the informational revolution and had to “re-invent” themselves as knowledge workers. Now, knowledge workers are being obsoleted by AI, and it is unclear how they they can reinvent themselves to survive. As someone once said: “If every job is a function of our intelligence, as AI beats us at intelligence, how could any job be safe?”. Put another way, in a human dominated world, smart people are in control; in a machine dominated world, nobody is smart enough to be in control. ALL humans are slaves.

But it is important to note that unlike automation, AI displaces jobs not by way of efficiency but by way of obsoleting business models (think how the iPhone completely obsoleted the taxi business. Nobody could have foreseen that at the launch of iPhone 1) and entire industries wholesale.

• In theory, AI best flourishes with abundant “permission-less and trustless” data distributed all over the world under the Web3 regime. Web three has its own challenges. First, there is the blockchain trilemma: Of speed, security and scalability, only two can be had at the sacrifice of the third. Then there is the dilemma of speed and decentralization: Faster networks often require more centralized control. Even after Web 3 becomes fully functional, if the (abundant) data is inaccurate, incomplete, or inconsistent — by intention or not — AI can easily perpetrate falsehood convincingly — resulting in misdirected, rapid actions in mass, likely irreversible.
• Fiscally, head-scratching on how to “tax AI” to make up for lost tax revenue from jobs displaced is pointless. Government will simply massively increase subsidy (read: Universal Basic Income) to support the unemployed. This subsidy can only be funded by government deficit spending (more money printing), worsening an already shrinking real economy. Worse, the unemployed on subsidy will lose independence and be enslaved (via a government-controlled digital monetary system such as CBDC).
• The world is already insolvent from over-indebtedness. The cash flow to service and repay debt comes from enterprises big and small in the economy. But the exponential growth in productivity eliminates a lot of activities in the economy, thus reducing cash flow. So then government has to issue more debt, making an insolvent world more so, — until it breaks.

I summarize my view of human progression pictorially thus:

For all the wonders we attributed to technological advances in the “Information Age” of the 20th century, they turn out to be mere building blocks for true “shock and awe” unfolding in the 21st century.

Rapid adoption of a harmless technology may be a good thing, but that of a potentially dangerous technology is very alarming. We already saw the exponential networking effects in social networking. We now know that social networking has life altering — if not outright life saving/destroying — power on massive scale. The un-subscription by celebrated thought leaders who previously embraced this technology attests to the recognition of its potential danger. ChatGPT’s user base grew from 1M in Dec 2022 to 100M in Feb 2023. That’s two orders of magnitude growth in 2 months! This eye-popping adoption rate — far exceeding early adoption of Facebook — is fast tracking generative AI adoption in general. Unfortunately, un-subscription from mass-deployed AI (via Open AI)— portending far greater impact than social networking — may not be an option. And then, dystopia is born.

There — I set the table with a “We’re all going to die” thesis as a “devil’s advocate” invitation for readers of this article to come back with a contrarian “I am wildly optimistic” thesis, possibly along the lines of:

• Networks erode boundaries and allow quantum leaps in productivity. Generative AI will put networks on steroids and bring exponential — not linear and incremental — growth to productivity. Communities (Metaverses, if you will) built on hyper-charged networks will further erode institutional power (already weakened as a hallmark of all 4th Turnings) and empower the individual. In turn, capital will follow “causes” rather than power structures.
• Open source AI will deny centralization of Generative AI’s power by Microsoft or Google, and instead enrich distributed Metaverses (which didn’t die with Meta but will probably be launched by a much more capable Apple). Creative digital citizens no longer content being corporate slaves will selectively opt-in these rich trans-border digital communities and collaborate on compelling Generative AI user cases for good and not evil.
• Between now and “we’re all going to die!”, many companies will build customized models for their unique user cases that are tremendous productivity picker-uppers but harmless to the world. For instance, some company may parse decades of internal emails (big data) and dispense expertise in building its products much more effectively than “expert systems” in the past.
• “Everyone can code” (as well as design, compose, present etc.) with ChatGPT4, removing dependency on specialists. Thus, a lot more ideas can be “productized” and “democratized”, — and a lot quicker. The “blooming of a million flowers” may wind up being a great boost to democracy and deterrent against autocracy.

Will the next First Turning be the beginning of Renaissance 2.0 or 1984? A large part of the answer lies in whether generative AI and companion technologies help centralize, or distribute power. The other part lies in whether digital communities can self-police.