We’ve now had an entire year for the mainstream populace to play with and grasp the consequences of a superintelligent AGI — namely, ChatGPT (and its siblings, children, and distant cousins). Now, it’s time to look at our AI forecast, and what that bodes for us and our civilization. For ease of digestion, I’ve broken this down into 6 epochs: hypercapitalism, foom!, social unrest, UBI, the Merge, and post-humanism. Buckle your seatbelts, fellow humans; we’re in for a wild ride!
AI Forecast: Enter Hypercapitalism
The first wave of AI follows the direction and policies of its masters: namely, powerful multi-national corporations, plus a handful of proxies (OpenAI for Microsoft ($13B — Apr 2023), Anthropic for Amazon ($4B — Sep 2023), Character.ai for Google ($0.4B — Nov 2023), etc.).
We already have hard evidence showing that the Fortune 500 is deploying AI in production environments at breakneck speeds — while ChatGPT (3.5) was the proof-of-concept, GPT4 is being deployed to the front lines — and as a result, laying off tens of thousands of entry-level and “repetitive task” employees whose jobs it is rapidly replacing.
Our AI forecast sees the same thing happening in creative corners — anecdotal stories of renowned illustrators who have spent a lifetime honing a signature style and personal craftsmanship — whose “uniqueness” has now been jacked by an AI bot digesting hundreds of internet postings of their work, and which can now spit out new product “in the style of (insert artist name here)” in under 60 seconds, without even breaking a sweat (and for pennies!)
It is of note that these are by and large white-collar workers. While it was assumed, long ago, that the first wave of AI would displace (the political term — job displacement. translation: replace humans with AI software and AI-powered robots) blue collar factory workers and truck drivers, navigating the gravity and dust and dirt and mess of the physical world has proven illusive. Much easier for AI to optimize the dataflows of information — white collar — worker bees.
So: who does all this benefit? First and foremost, rapid AI deployment grants a pronounced competitive advantage to those who deploy it first. Which leads to three results: greater profitability, increased market share, and as a result, increased stock price and market capitalization. But since the likely winners of the AI race are already some of the largest corporations on planet earth, we can reduce this to: the biggest get bigger.
Despite the claims of the technorati, our AI forecast does not see the benefit trickling down to the workers (other than those in Silicon Valley with stock options). No, all this optimization benefits primarily two entities: the executive staff of the multinationals, who claim the performance bonuses, and the investors (both institutional and consumer) who have both the savvy and the financial wherewithal to have invested in the winners.
So, in short: expect to see a dozen or so global winners, across key industries (telecom, energy, healthcare, automotive, tech, etc.) skyrocket into $10 trillion valuations (for comparison, Apple & Microsoft are currently the only AI-forward companies with market caps north of a trillion USD)
AutoCorps: the autonomous AI corporation
There is, of course, a dark horse in this race. (There’s always a dark horse, isn’t there?). And that dark horse is something that has been well explored in the science fiction canon, but only touched on lightly by those crazy enough, brilliant enough, or secure enough to mention it: the rise of the fully autonomous, zero-human, 100% AI corporation. Our artificial intelligence AI forecast plots the probability of this entity as… 100% within 3 years.
The US Supreme Court created the idea of corporate personhood in 1886 (Santa Clara County v. Southern Pacific Railroad), which essentially gave, in broad areas, corporations the rights of citizens… of humans. And in case after case since then, the courts have re-affirmed that position.
I’m fairly sure that in 1886 the justices had no inkling of a corporation 100% managed, operated and run by machines… but in 2023, this is a completely realisable scenario, with no real practical or technical issues barring its implementation (who is to say, in fact, that it hasn’t already occured?)
In fact, for many years I’ve harbored this simple fantasy: seed a bank account with $100,000 USD, give the keys to an AI, and instruct it to automatically form as many inter-linked LLCs and corporations as necessary, to employ as many sub-AIs as necessary, all with the goal of maximizing revenue and profit growth, using whatever means it sees fit.
Sutskever on AutoCorps
Our AI forecast is not alone in this view. None other than Ilya Sutskever, co-founder and chief scientist at OpenAI, waxes poetic:
“The way I think about the AI of the future is not as someone as smart as you or as smart as me, but as an automated organization that does science and engineering and development and manufacturing…” Suppose OpenAI… builds and AI with a rich conceptual model of the world, an awareness of its immediate surroundings, and an ability to take action in the physical world… not just with one robot body, but with hundreds, or thousands, working in concert towards a common goal…
“We’re not talking about GPT-x. We’re talking about a fully autonomous corporation.”
Its constituent AIs would work together and communicate at chip-speed, like bees in a hive. A single such AI corporation could easily be as powerful as 50 Apples or Googles, he mused. “This is incredible, tremendous, unbelievably disruptive power.”
— Ilya Sutskever, interviewed by Ross Anderson
“Inside the Revolution at OpenAI”
The Atlantic, Sep 2023
These “activities” would include grey-area possibly illegal actions, as long as the calculus decided that the punishment would be less than the profit (can you say tobacco industry 1998, banking industry 2008, Apple’s ongoing IP-licensing tax evasion?). This entity, of course, requires the benefits of a sustainable supply: not only in terms of raw material, but also in terms of energy and compute.
An AI corporation would have none of the delays of human verbal communication, or the need for sleep. It wouldn’t have sexual harassment lawsuits (unless those enhanced its profitability). It would be hyper-efficient at pursuing its goals. It would be… hypercapitalism, at its apogee. And then…
and then… foom!
At some point, as you might imagine, our AI forecast sees this hypercapitalist ascent as becoming unsustainable, both financially and ideologically for the “normal” citizens of planet earth. At about the same time, the AIs are not slowing their performance gains… in fact, they are accelerating.
And so our AI forecast enters its most delicate moment: fast takeoff, or more dramatically referred to in certain circles as… foom! (as in, the last sound you hear as all the oxygen is sucked out of the room right before a massive fireball explosion… not to put too negative a spin on it… just sayin.)
In my humble opinion, the tipping point occurs when som corporation aims its substantial intelligence at the heart of the beast: the global financial markets. While fully automated algorithmic trading has been a thing now for more than 3 decades — we can thank Jim Simons and Renaissance, the top performing hedge fund in the world, for that insight.
And when some superintelligent AGI is handed a $100 billion dollar fund to manage, and starts driving 3,000% year-over-year returns, and then a couple competitors join in… well, I fully expect the market to go completely sideways at that point, tripping the NYSE and NASDAQ circuit breakers 1,000x per day, as opposed to the present brisk pace of .
I’m not saying, btw, that our AI forecast predicts a market crash. Honestly, I’d put it at 50/50 whether the market goes up or down once the AIs take hold. They don’t care if they’re short-ing or long-ing stocks & markets — they win with volatility and accurate prediction. But when market moves happen in under a millisecond and in trillions, as far as human perception is concerned, it starts getting very… noisy.
As AI deployment continues, and layoffs increase, our AI forecast predicts global unemployment skyrocketing, perhaps hitting as high as 35% by 2030. If the stock market doesn’t send a clear signal, this certainly will.
And with unemployment of that magnitude, as well as the simultaneous hyper-concentration of wealth that hyper-capitalism creates, civil unrest is almost a newar certainty. And again, at some point, the pressures get too extreme, and something has to give.
In the optimist scenario (which we always advocate), the re-organization & re-distribution of global wealth is centered around the concept of UBI: Universal Basic Income. The measurable benefits of a sustainable economy, for humans and AI together, cannot be understated. (AI wrote that line for me. The only line in this article. I despise it, but hey, it just bumped me 11 places in the search rankings. Heaven help us all.)
UBI: from where?
Three possible methods, from best to worst:
- OpenAI
and its capped profit model: spawns a few thousand autocorps, solves the stock market, pays off its investors, converts to a global non-profit, and pumps a few trillion USD into WorldCoin. - The United Nations
asserts actual authority, and in cooperation with the old world institutions of the IMF and World Bank (with commercial partners Wells Fargo, Goldman Sachs, and Bank of America), figures out a way to do UBI. (our AI forecast sees this inevitably using a novel cryptocurrency to manage this, so it may be closer to option 1 than it initially appears on the surface.) - The People
exert their will, forcibly. The French Revolution, replayed in the 21st century. Not pretty.
Accelerando: REquired REading
Luckily, all this has already been envisioned (way back in 2005, no less!) in a highly entertaining (and mildly disturbing) narrative by Charles Stross (aka antipope). A simple google of the title will find the book for you. In a prescient move, Stross made the manuscript open source, so you can find it as a free download in just about any e-book format you can imagine, as well as in charming turn-of-the-century HTML 2.0.
In some ways, our AI forecast is simply a practical manifestation of the broad cultural trends that Stross, a true visionary, imagined in his epic novel. The actions you take based on this intel, are, 100%, your business.
keywords:
- AI forecast
- Hypercapitalism
- Automonous Corporation
- AutoCorps
- fast takeoff
- wealth redistribution
- UBI – universal basic income
- Accelerando