We Are Not Ready For AI

There is a major externalization caused by AI-based technology for which modern capitalism and the tax-reliant state is highly unprepared.

Artificial intelligence (AI) will allow further accumulation of wealth in the hands of a few and removing the source of wealth (i.e. jobs) from communities that sustain people.

The Last 30 Years Are a Cakewalk Compared to What Will Happen In an AI-Driven World

First it was box stores taking out the family business on Main Street. Capitalism at work.

Then Internet retailers like Amazon began taking out box stores (book stores first and now Toys ‘R Us and potentially even Wal-Mart). More loosely-controlled capitalism at work.

Here’s what will happen next:

AI will take out knowledge workers. This will not be more loosely-controlled capitalism at work the way we’ve experienced it before. AI will destroy white collar jobs, which have essentially been immune to prior destructive cycles caused by new technology. This AI takeover will be supported by “news” that an algorithm has written or compiled for you to closely match your emotional needs. You will BELIEVE it is the right thing because you, as a good white collar worker, will predictably ascribe to capitalistic theory and consider it economic “advancement” or “renewal”. But you will suffer.

The move to efficient technology (example: AI for Customer Support) will drive lower prices, which is theoretically good for consumers. But it will also remove the ability of consumers to pay when they lose their jobs or experience real salary erosion. Since AI will primarily eliminate white collar jobs along with their high salaries, this will erode the middle class purchasing power and tax base that drives industrialized economies.

AI will allow large tech players like Amazon, Google, Facebook and Apple to accumulate wealth at an ever-accelerating pace. Their shareholders and employees will benefit. The industries they absorb will essentially be deconstructed, with AI taking over and no possibility of competing with less-efficient means.

It cascades further.

The state will likely collapse. With fewer local taxpayers municipalities lose tax revenue and the ability to fund services like education. The pace of change and lack of trickle down will simply get too far ahead and there will be no recovery possible for tax-based towns, cities, states and nations.

We will need to develop an entirely new economic system to force recycle capital from the wealth accumulators who use AI like Amazon and Facebook (and the next uber tech companies) or we will quickly become the Untouchables and the Destitute.

“Just move or retrain yourself then.” (Why the Personal Responsibility Argument Fails)

The worker/consumer can’t just move and find another job because the AI takeover will cause a systemic elimination of jobs across all geographies.

The worker/consumer also can’t just switch to an AI tech job because that requires retraining and up-front costs, or they simply won’t be able to learn the new skills.

Healthy systems require some destruction of weak or inefficient actors. But when the destruction occurs extremely fast it doesn’t lead to renewal. It leads to death.

Look at the Rust Belt and you can see what efficient technology-driven global supply chains leave behind when factory workers have no viable employment alternatives. Poverty.

How Close Are We To “AI-megeddon”?

With AI growing so fast I see this process accelerating.

We are in the beginning stages of widespread AI use. Bots used for things like answering questions online and navigating through phone responses are extremely simplistic versions of what AI can do when combined with Natural Language Processing (NLP) and allowed to self-train.

Instead of hitting 5 for “cable box installation” and navigating through pre-created menu items before finally speaking with a live agent, an AI agent using NLP will simply ask “how may I help you”. You will be assisted by a machine who is essentially an uber-human with access to all the knowledge available on the subject matter. This one AI agent will be infinitely scalable, handling millions of support calls without a single human agent necessary.

Tax Accountants beware, your job is next. An AI agent will quickly learn all local, state, federal and international tax laws, and every use case these rules apply to, and spit out answers thousands of times faster than a human CPA.

Or maybe truck drivers are next. Taxi drivers are already looking over their shoulders as an army of Google cars is poised to eliminate the need for human hands at the controls (or controls at all).

The “lanes” that AI agents are designed to operate in can be as narrow or wide as their creator devises. While there may be different applications, the capacity of AI plus NLP to handle a much broader range of problem-solution sets means the creation of a super-AI entity is highly possible. An AI agent given extremely wide (or no) “lanes” to operate in will learn so much so fast that it will become very powerful in a very short time. If this becomes reality then no jobs are safe from AI (including the Creator of the AI agent itself).

The “You’re An Idiot” Counterpoint

You may say:

“The same wealth creation happened with railroads, industrialization, and agriculture. And each time life got better for everyone. Overall wealth, lifespan, infant mortality, everything got better. This AI thing is good. Don’t be an idiot!”

You may also say:

“Individual responsibility to find new, higher value employment is the solution to job loss. If you don’t let innovations like AI happen then you’re promoting Neo Socialism, and Neo Socialism is Socialism — a totalitarian mindset, top down, exclusive, authoritarian. It will take away incentives to wealth, and cause the middle class to vaporize and you will get the nihilists on the bottom and the Uber elite at the top. It happened in The USSR and China. You can have it too. And it will be very bloody.”

Hmmm…AM I being an idiot about this AI thing? Could AI actually be the solution to the world’s ills the way industrialization of Western society surely was?

I mean, nothing BAD ever happened from industrialization, right? (Don’t answer that, it will take us way off-topic.)

Here’s the problem with using industrialization arguments:

AI Is Not Industrialization!

Artificial Intelligence is not industrialization. The purpose of industrialization is not to eliminate human jobs — it is to create more output for each unit of human work. Job elimination is an effect of industrialization, but so is job creation. The purpose of AI, on the other hand, is specifically to eliminate human jobs.

Eliminating one type of job, say manual wheat farming, and replacing that with humans driving tractors has the purpose of increasing available food output for the population. Tractor driver jobs are created from the destruction of manual farming jobs. Replacing farmers entirely with AI would eliminate humans from the agriculture equation entirely and concentrate all wealth in the hands of the agriculture AI creator / controller.

Similarly, providing a doctor with a robotic arm that allows her to do brain surgery far more quickly and accurately has the purpose of increasing medical effectiveness, speeding up surgeries and saving more patients. Robot designer jobs are created. Replacing the doctor entirely with AI would eliminate the human from the surgery equation and concentrate all wealth in the hands of the hospital (the AI creator / controller, presumably).

Here’s the catch:

There is a fuzzy line between “good” job elimination and “bad” job elimination. Nobody in their right mind is arguing that we should revert to a subsistence living where everything is manual and there is minimal specialization. However, when we look at the potential — and imminent — power of AI, we can conceive that massive numbers of high value jobs might be eliminated overnight by this technology, with no jobs to replace them. Jobs will simply evaporate, leaving a major void.

This is scary because our entire economic structure is based on employment and the exchange of wealth from that employment to obtain goods and services we need. We have no viable system for a mass-unemployed world. The merry-go-round just stops.

Looking back will not help us. The “individual accountability / find a new place in the world” argument is seriously flawed because personal initiative will not help much when AI takes over.

Left unfettered AI will do this:

1) Eliminate most “jobs” as we know them. CPA, doctor, truck driver, clerk, M&A advisor, etc. “Jobs” will evaporate. Individuals will need a different solution than a “job” to sustain themselves economically. With industrialization a cobbler could switch to a factory job fairly readily. Not so with AI.

2) Eliminate the individual tax base required to fund common services that allow freedom and industrialization to thrive: highways, courthouses, air traffic control, drug enforcement, national defense, university R&D, shareholder laws. No money no services. The pieces that create healthy competition like anti-trust will go away for lack of funding.

3) Set us on a highly unpredictable survival path. In the best case scenario we will end up with abundant and virtually free food, shelter, education, healthcare, entertainment, etc. with some system of chits to exchange that constantly gets refilled. Another possibility is uber-corporations/shareholders that replace the nation-state and mete out sustenance to a chosen group (New USSR). In the worst case the entire economic system collapses and we all end up back in barter stage.

It’s Going to Be a Wild Ride

I am in no way a socialist. I’m an Ayn Rand fan generally. I believe human-human competition is essential to community health. However human-AI competition has massive potential to erode health and welfare over time and we really must consider controlling it. Or AI will quickly control us.

This AI thing is in motion and we’re not prepared. The old solutions won’t work in an AI-driven world. I do believe we’re in for a major upheaval and reorganization of our political and economic system.