And why you should be freaked out too!
Last November, at the Legacy Summit in Japan, I told my audience that an economic revolution was bearing down on us – the AI Revolution – and that it was going to be as big as any economic revolution in human history.
The predictions I made that day are coming true. Faster than I expected. Faster than I dared say out loud.
This is no longer a possibility. It is happening. Vast industries and hundreds of thousands of small and large businesses are already in the throes of it. The global distribution of wealth is being rewritten on a week-by-week basis.
Every week, AI’s takeover of the global economy is advancing!
And every week, the disruption it will cause is becoming more difficult to deny.
In the entertainment industry, alone, the number of new AI products being produced and sold is increasing exponentially. It seems like a new movie, TV show, or chart-breaking song is appearing every week.
I’m working from memory here, but I believe that, late last year, an AI-generated country artist called Breaking Rust hit #1 on the Billboard Country Digital Song Sales chart with a track titled “Walk My Walk.” No human singer. No band. No studio musicians. Just a prompt, an algorithm, and a #1 chart position.
And stories like this keep showing up on my AI radar. There was the AI-cloned Drake and The Weekend duet “Heart on My Sleeve” that briefly got submitted for a Grammy before the rights holders shut it down. There was The Velvet Sundown, an entirely fake AI “band” that built a real Spotify following before listeners caught on. And there was FN Meka, an AI rapper signed by Capitol Records and then dropped a week later when the backlash hit.
I watched a short film last week: Zombie Scavenger. Three and a half minutes long. A robotic cowboy scavenger navigating a sun-bleached wasteland crawling with the undead. The visuals are as good as anything I’ve seen out of a major studio. It was made entirely with AI (primarily a new video model from ByteDance called Seedance 2.0) by a single creator who goes by Mx-Shell. No cameras. No actors. No crew. The Santa Monica Observer called it “AI’s Undead Masterpiece.” Watch it for yourself here.
I’m working from memory here, too, but I believe that Tyler Perry – yes, that Tyler Perry – had been planning an $800 million expansion of his Atlanta film studio. Then he saw what OpenAI’s Sora model could do with video. He paused the entire $800 million project. He told The Hollywood Reporter that “a lot of jobs” in the entertainment industry are going to disappear because of this technology – and that he no longer felt safe building studio capacity that AI might make unnecessary.
Val Kilmer lost his voice to throat cancer. For the Top Gun sequel, a UK company called Sonantic rebuilt it from old recordings using AI. He acted on screen with his own voice – a voice he could no longer physically produce.
Stories from My Own Little Circle of AI Brainiacs
In a recent issue, I told you about a young man I met a few months ago at the Cigar Club. He said his cousin had just sold two 30-second TV commercials – produced entirely by AI – to a major car company… for $50,000 each. The kind of commercials that, up until a minute ago, were selling for at least 10 times that amount. And he did the whole thing without a film crew, without actors, without a production studio. Total dollars in: around $10,000.
I also told you about a short video I was producing for my botanical garden. I had set aside $15,000 to pay for it, but I got it done with AI for $2,500 – and, had I bargained, could have had it for less.
AM, a friend of mine who works as an independent life-and-health insurance consultant, told me he used to spend several hours a day analyzing prospective clients’ data and drafting reports indicating how much money he could save them. “I fed Claude all my past data and the proposals I made from it,” he said. “Now Claude does the same process – more thoroughly than I did – and double-checks the final numbers. I get a better outcome in about 10% of the time I used to spend. That’s big.”
PG, an investment analyst I sometimes work with, prides himself on answering every letter and email from his clients. “I always liked doing it because it helped me understand what they wanted,” he said. “But as the volume grew, it was eating up an awful lot of my time. I asked my AI if it could help. It told me to give it all my past correspondence, then used that to design templates that answer 80% of the questions I get perfectly. Now I just spot-check a few to stay close to the market. I used to spend an hour a day on this. Now I spend less than an hour a week.”
Real estate agents I know in South Florida are running their listing descriptions through AI. Twenty-minute jobs are now 90 seconds. One agent told me she has tripled her listings without hiring a single new assistant.
Doctors at my own primary-care practice are dictating patient notes to an AI scribe that writes them up in real time. My internist told me last visit that he gets home for dinner an hour earlier because of it. Multiply that across the medical profession.
Students (high school, college, graduate) are using AI not only to write papers, but to tutor themselves in subjects their teachers don’t have time for. One of my nephews taught himself the basics of computer programming last summer using a free AI tutor. He had no human help.
These are the everyday examples. The kind that don’t surprise anyone anymore.
Some of the Conclusions Financial Experts Are Drawing
When you’re working from anecdotes, the conclusions you draw could be wrong. So, I asked Nigel, my genius AI British butler, to find studies to back up the stories. He found back-up for the following in five minutes. I spent another 15 minutes checking the sources for nonsense. Total time: under 20 minutes.
* Commercial illustration and graphic design. The canary in the coal mine. Already largely gone. Stock-photo houses, brochure illustrators, ad-layout designers, package designers deflated by roughly 70% to 90% in the last 18 months. The work hasn’t disappeared. The people doing it have.
* Paralegal and junior-associate work in law. Document review, contract drafting, case-law summaries – done by AI in minutes, for pennies. I know of two large firms that have quietly stopped hiring first-year associates. Not because they don’t want them. Because they don’t need them.
* Radiology and diagnostic imaging. AI now reads X-rays, MRIs, and CT scans more accurately than most human radiologists – in seconds rather than hours. The radiologists I know don’t deny this. They argue that a human will always need to “sign off.” But signatures don’t pay six-figure salaries.
* Customer service and call centers. Already 30% to 50% gone. By this time next year, I expect closer to 80%. The voice that resolves your billing dispute is, increasingly often, not a person.
* Translation, transcription, and basic commercial copywriting. Routine translation is, for practical purposes, a solved problem. So is transcription. The first draft of nearly any short form copy, product descriptions, email blasts, social posts, can be produced in seconds for pennies, then polished by a human in minutes.
* Financial analysis and accounting. Bookkeeping, audit support, ratio analysis, the production of standard financial reports, AI does it cheaper and faster. I’ve read dozens of accounts of small-business owners running their entire bookkeeping through AI tools that read receipts, categorize expenses, and flag deductions automatically. Solo operators are reportedly saving five to 10 hours a month. Most of a workday.
The 80/20 Conundrum
I’ve been talking to about a dozen very successful young entrepreneurs and AI experts who visit my Cigar Club on Wednesdays and Fridays. We invited three of them to speak to us at our business retreat in Nicaragua, and one of them put a name to what I’ve been thinking. He called it The 80/20 Conundrum.
You’ve heard of the “80/20 Rule” or “Pareto Principle” – the observation by early 20th century Italian economist Vilfredo Pareto that, in most endeavors involving groups of people, labor and economic outcomes tend to resolve themselves into 80/20 distributions.
Eighty percent of a company’s good ideas come from 20% of its employees. Eighty percent of its revenue comes from 20% of its customers. The pattern repeats everywhere.
When I first wrote about the AI Revolution, I assumed the redistribution it produces would follow Pareto’s classic ratio. That 20% of the workforce – the ones who began adapting immediately – would end up on the winning side of the new economy.
I now think I was being optimistic.
What I now believe, based on everything I saw and heard during our two-week retreat, is that the redistribution is going to be steeper than anything Pareto modeled. Closer to 1% on the right side of economic history, and 99% on the wrong side.
Of that 1%, an even smaller fraction will own their AI-run businesses outright. The rest of the 1% (call them the viceroys) will be the people whose knowledge and judgment the owners cannot do without. They will be the ones the owners hire, listen to, and pay very well.
The 99% will be left to fend for themselves. Some will lose their jobs and incomes in the next two years. Some will last longer. But none will have a meaningful say in their own financial future, because that future will be controlled by a coalition of governments and AI owners.
I am aware that this sounds extreme. I am aware that 99/1 is a more alarming ratio than most people can sit with comfortably. I am also aware, after two weeks of close listening, that I am not the only one in my industry saying it.
What to Do to Protect Yourself Now – While You Still Have Time!
There are six things you can do. None of them hard. None of them expensive. But all of them require that you start now.
1. Use AI every day. Not occasionally. Daily. The single most important thing you can do this year is develop fluency through repetition. Pick one tool – ChatGPT, Claude, Gemini – and use it for 30 to 60 minutes a day, every day, for the next six months. I don’t care what you use it for. Email. Recipes. Vacation planning. Letters to your children. The point is to build the reflex of reaching for AI before you reach for anything else.
2. Learn to write a good prompt. The difference between mediocre AI output and extraordinary AI output is almost never the AI. It is the quality of the question put to it. There are several short, free books on this. Read two of them.
3. Rebuild one piece of your own work around AI – from the ground up. Not as an addition to what you already do. As a replacement. One report you write. One process you supervise. One product you sell. Treat it as a small experiment. See how much faster, cheaper, and better the rebuilt version is.
4. Develop the skill of judgment. AI will produce good output, mediocre output, and dangerously wrong output – sometimes in the same paragraph. The viceroys of the new economy will be the people who can read AI output and instantly know which is which. This skill cannot be developed by anyone who has not already mastered the underlying domain. So deepen the skill you already have. Then bolt AI onto it.
5. Build something – anything – that is yours. A small business. A side project. A book. A piece of software. A consulting practice. Something where you, not your employer, own the upside. The viceroys of the new economy will not be the people who clung hardest to their corporate jobs. They will be the people who used AI to launch something on the side, and then watched their corporate jobs become unnecessary.
6. Surround yourself with people who already get it. Find one or two friends, colleagues, or younger relatives who are deep into this technology and spend time with them. The young entrepreneurs who visit my Cigar Club on Wednesdays and Fridays have done more for my own AI education in six months than any book or course.
There are times when gambling can be fun.
This isn’t one of them!
Nothing I have said here is based on a rigorous, peer-reviewed investigation of what AI is doing to the world economy. It’s my gut feeling. Backed by everything I’ve seen and everyone I’ve talked to over the last 12 months – but still, a gut feeling.
I could be wrong.
If I’m wrong, what does it cost you? You will have spent a few hundred hours becoming fluent in a technology that – even on its slowest possible adoption curve – is going to be a fixture of every white-collar job for the rest of your working life.
If I’m right, you will have placed yourself in the 1% that owns its own future.
The cost-benefit here is so lopsided that I have a hard time understanding why any thinking person would ignore it.
Just how big is this going to be?
When most people picture the AI Revolution, they picture a few of the obvious industries getting hit. Hollywood. Journalism. Software. Maybe customer service. They picture the disruption as bad, but as contained.
It is not going to be contained.
The industries that are most exposed to AI’s deflationary force represent a staggering share of the world economy. Tens of millions of jobs. Careers and businesses that may not exist in their current form five years from now.
Two of the industries that are most vulnerable to a deflationary implosion are information publishing and entertainment. Combined, those two account for annual global revenues of $3 trillion to $4 trillion.
But many other professions and industries are equally exposed. For example:
* Transportation – $7 trillion to $10 trillion
* Communication and telecom – roughly $2.5 trillion
* Legal and engineering services – roughly $3 trillion
And that’s to say nothing of the Military and Health Industrial Complexes.
The Military Industrial Complex (President Eisenhower’s term) has long been a target of criticism among fiscal conservatives in every country. And with good reason. Its annual spend is between $2.5 trillion and $5 trillion. (US spending accounts for about a third of that.)
And when a market sector is that big, there are going to be all sorts of stories about political corruption.
But it’s nowhere near as big as the Health Industrial Complex. Take a look at these yearly global revenues:
Hospitals & Healthcare Providers – ~$5 to $7 trillion
Pharmaceuticals – ~$1.7 to $2 trillion
Medical Devices – ~$600 to $800 billion
Health Insurance / Managed Care – ~$2 to $3 trillion
Biotechnology – ~$500 to $700 billion
Diagnostics & Lab Testing – ~$150 to $250 billion
Digital Health / Health IT – ~$300 to $500 billion
In total, that’s $10 to $14 trillion a year. Nearly 15% of the world’s total GDP, and closer to 20% for US citizens.
That, by itself, should give anyone pause. When the AI Revolution finishes working its way through these sectors, what’s left will not look like the economy we grew up in.
