In 1925, a Nashville insurance company put a radio station on its fifth floor to sell policies. Its barn-dance show became the Grand Ole Opry. 

It’s still on the air. 

That same decade, RCA stock rose 200-fold. 

Then it fell 98%.

Radio changed the world. 

Radio stock wiped out everyone who bought the top.

This week, a century later, our whole team came to Nashville for our annual Paradigm Shift Summit. One question in the room: Is AI the new radio?

The answers I’ve gathered this morning reveal where the money’s moving, when to buy, what would mark the top… and how to play it, in one sentence. 

A Show of Hands 

JC Parets started by taking the room’s temperature. He asked investors to raise their hands if they use AI every day.

Hands went up.

Then he narrowed it. Who uses it to write emails and run their calendar? Who uses it to write code and build software?

Some hands came down. 

"Barely anybody is using AI for real," JC said. "A lot of people are using AI like a glorified Google search."

He cited a stat: only 2% of American households pay for AI.

JC is one of the 2%. That morning, his AI pulled the weekly investor sentiment survey before he remembered it was Thursday. It also sent his Uber to the airport and kept him posted on gate changes. 

That's the real great divide. On one side are the people building with AI. On the other are the people reading magazine covers about it.

As with all industries, the money follows the builders, and the people paying close attention to them.

Greg Guenthner and Nick Riso built a new stock scanner. It has real-time pricing, custom filters, built-in charting and a sector-flow view. A few years ago, this would have taken an entire development team and at least tens of thousands of dollars. Now it takes one developer and AI. 

Ray Blanco talked about a team that used Claude to design scorpion anti-venom candidates in 64 hours. Also, Revel Pharmaceuticals and Google’s Calico used AI to engineer an enzyme that cleans sugar damage out of old tissue. In a 75-year-old donor’s aorta, it rolled the damage back to the level of a 30-year-old.

The Bottleneck Moved

When ideas get cheap, everything they need gets expensive.

That’s where the money flows. The best investors catch them early. 

The anti-venom team found out fast. Design took 64 hours. Lab testing took months and far more money. Verification and development—where one company called Bruker (BRKR) sits—is the growing toll booth. 

Byron King covered the second bottleneck: atoms. He held up a piece of a turbine blade from the engine of a B-2 bomber, the most secret airplane in the world. (Where’d he get it? eBay.)

That blade survives the heat because of a few percent of rhenium. Rhenium comes out of molybdenum, which comes out of copper. If the smelter doesn’t catch it, it goes up the smokestack and disappears. Space, defense and AI all run on obscure metals like rhenium. China and Russia dominate much of the refining.

You can’t open a rhenium mine. Supply depends on how much copper gets dug up, and whether anyone bothers to catch the rhenium on its way up the stack. (THAT, says Byron, is the opportunity.) 

Then Enrique Abeyta named a third bottleneck: power. Amazon has roughly $100 billion of chips it can’t use. The racks are built. The chips are stacked. Nobody can get electricity to them.

Development. Atoms. Power. The ideas are almost free. Everything after the idea costs a fortune—and that’s where the money is moving.

The Big Question: Timing

Every speaker brought a clock. No two clocks agreed.

Now. JC says this is the best window of the four-year presidential cycle. Over the past 100 years, the Dow has gained about 50% on average from the low of a midterm year to the high of the following year. JC says we're up 16%. His picks include Pulte Homes (PHM) above $125, with a $200 target, and Newmont (NEM) above $110.

The next six months. Enrique agrees with JC. Then he gets off the bus. Since ChatGPT launched, Nvidia is up 1,109%. His three-year call: the Nasdaq down 60% to 80%, the S&P down 50% and small caps down 80% to 90%.

He gave three warning signs. First, the Fed raises rates. Second, Nvidia's earnings estimates get cut. Third, the 50-day moving average crosses below the 100-day and 200-day. He expects the first two in the first half of next year.

December. Byron says wait. Tax-loss selling should make good companies cheaper over the next month. Buy on down days. Use limit orders. He thinks gold bottomed last week.

Sit out? Sean Ring opened the day by confessing his silver trade. On January 30, silver fell from $120 to $78 in one candle. His advice now: it's all right to sit out. And use bracket orders, so the stop is set before your emotions show up.

The Overlap

The bull and the bear agreed on one trade.

JC says bonds are ripping and rates are heading down. Enrique says bonds will be the one thing that goes up. Different reasons. Same trade.

Two Things Can Be True

Enrique gave the morning its thesis. The market could be great over the next six months and great over the next six years, he said, "and in between you might lose everything."

So here's the plan in one sentence: 

Ride JC’s window, buy Byron’s bottlenecks on down days, follow Greg’s trends, own Ray’s toll booth, set Sean’s stops and tape Enrique’s three signs to my monitor.

Keep Ray's line handy, too: "None of these ideas matter if you're dead."

Be right about AI. And be around to collect. 

Tomorrow, we’ll cover the afternoon: a 100-bagger chip, the future of AI, a debate between James and Jim, and the famous Whiskey Bar.