Two of the loudest financial publishers in America are running directly at each other this summer, and the audience they are fighting over is the same person.
On one side, Jim Rickards and Porter Stansberry are campaigning an AI debt thesis. AI is a credit bubble, they say. Hidden leverage, off-balance-sheet vehicles, $570 billion in AI debt issued in 2026 alone, and a July 29 catalyst that Rickards calls a potential dotcom moment. The Dow drops 80 percent in his framing. Porter published a May 2026 essay calling it the greatest financial bubble in history forming a top.
On the other side, Jeff Brown and Luke Lango are campaigning an AI supercycle thesis. Jeff Brown Anthropic IPO thesis frames 2026: Brown published 31 predictions on January 1 framing 2026 as the year artificial general intelligence arrives, GDP growth above 5 percent, and an IPO wave exceeding 2020. Lango is running the Genesis Mission thesis, $37 billion in annual federal AI spending across 26 Department of Energy challenges, with a six-sector playbook of companies positioned to capture it.
Both camps are spending heavily. Both are on the same email lists, the same Facebook feeds, the same conservative-news syndication circuits. Both are pointing at the same July 29 earnings window as the moment their thesis gets tested.
They cannot both be right. The interesting question is whether they can both be right in the way each one claims, and what history says about a market where two opposing macro theses are running at one audience simultaneously — which is the situation the dotcom crash of 1999 already answered once.
The Bear Camp
Rickards has been running the AI Black Paper presentation since March 2026, with 13 press releases through GlobeNewswire and Facebook scaling via conservative-news syndication. The thesis is specific: Meta earnings on July 29, 2026 could be the dotcom moment. AI companies have financed their infrastructure buildout with staggering debt — the core of his Jim Rickards AI debt warning, much of it off-balance-sheet, and when the credit cycle turns the structure collapses. He names an 80 percent Dow drop as the destination.
Porter published the credit-cycle essay on May 13, 2026, running across his Substack, the Porter & Co. member site, and The Burning Platform. His framing is sharper on the mechanism: the AI rally is a credit story, not an earnings story, and the credit cycle already turned. He points at CoreWeave taking out six-year loans against GPUs that depreciate in two, at CAPE ratios above 40, at $625 billion in data center debt. Whitney Tilson has been running a parallel Stansberry campaign around the same AI-doom framing for the same retiree audience.
The bear camp has independent support. JPMorgan built a credit default swap basket on five hyperscalers in February 2026. Morgan Stanley published research saying the AI supercycle ends if credit markets close. Michael Burry shared a scenario report in February captioned “And you think I’m bearish.” Ray Dalio said in June 2026 that his proprietary bubble indicators show AI at roughly 80 percent of the 1929 and 2000 peaks.
The Bull Camp
Brown’s 31 predictions, published New Year’s Day in The Bleeding Edge, frame 2026 as the year the productivity J-curve bends upward. He predicts GDP above 5 percent for at least two quarters, a Federal Reserve cut of 75 to 100 basis points, unemployment above 5 percent, and an IPO wave exceeding the $78.2 billion of 2020. The anchor call is that xAI achieves AGI by midyear on the back of its Memphis supercluster. The SpaceX IPO, which Brown predicted from his June 2026 Starbase visit, landed and briefly tied SpaceX with Amazon as the fifth most valuable company in the world within five trading days.
Lango’s Genesis Mission thesis, published through InvestorPlace, names $37 billion in annual Department of Energy AI spending across 26 challenges and six sectors. He tracks the equity injection playbook, where the government takes direct stakes in private AI infrastructure companies. His published portfolio names the federal free pick as MP Materials and maps the contract mechanisms — national lab operators, nuclear component manufacturers, semiconductor fabricators, quantum computing pure-plays.
The bull camp also has independent support. The hyperscalers have real cash flow, genuine pricing power, and dominant market positions. The Genesis Mission is a real executive order with real budget line items. NVIDIA hardware genuinely sits at the foundation of every major AI training run. Brown’s 2025 track record was tilted-correct on direction, conservative on magnitude — he predicted something close to AGI by end of 2025 and the industry broadly conceded that timeline, and he predicted the SpaceX IPO valuation exceed $2 trillion and the market confirmed it.
The Collision
Here is the part neither camp puts in their own promo.
Both theses can be structurally correct and still produce opposite investment outcomes, because they are arguing about different things on different clocks.
Rickards and Porter are arguing about the credit cycle. Their thesis is that debt service eventually crowds out productive spending, that refinancing risk hits a maturity wall in 2028, and that the repricing happens when credit markets get choosy. That is a structural argument about leverage and it does not require AI to fail as a technology. It requires AI debt to fail as a financing structure.
Brown and Lango are arguing about the technology cycle. Their thesis is that the productivity J-curve bends in 2026, that federal spending accelerates the buildout, and that AGI arrives by midyear. That is a structural argument about capability and it does not require the debt to be sustainable, because the technology landing is the separate question that the credit-cycle argument never engages. These are not contradictory claims; they are claims about different layers of the same stack, which means the technology can succeed while the financing structure fails and the financing structure can hold while the technology disappoints. The 1999 dotcom crash is the canonical example: the internet was real, the productivity gains were real, the companies that survived went on to define the next two decades, and roughly $5 trillion in market value still evaporated on the way there because the capital cycle and the technology cycle ran on different clocks.
What 1999 Actually Proves
The dotcom boom had two camps too. The bulls said the internet was changing everything and the bears said valuations had disconnected from reality, and both were right — the internet did change everything and the valuations had disconnected — but the Nasdaq fell 78 percent anyway, took 15 years to recover, and the companies that survived the washout — Amazon, Google, Apple — became the largest companies on earth.
The lesson is uncomfortable for both camps, because the bears were right about the crash and wrong about the technology while the bulls were right about the technology and wrong about the calendar. Anyone who acted on only one side of that collision either missed the recovery or rode the crash down. The investors who navigated it successfully understood that a real technology and an overpriced market could coexist, and they sized their positions so that a capital-cycle resolution arriving first would not wipe them out before the technology-cycle resolution arrived second. That is the framework both of these promo campaigns are implicitly arguing about, and neither one will say it out loud because the promo format needs you to pick a side.
Where This Leaves You
The two-camp collision is a risk to manage by holding both claims at once, not a puzzle to solve by picking a side.
The bear thesis tells you to watch leverage, refinancing walls, and credit spreads, and the bull thesis tells you to watch infrastructure buildout, federal spending, and model capability benchmarks. Both data sets are publicly available and both update weekly, which means the investor who wants to hold both claims at once can actually track both sides of the collision in real time rather than waiting for a guru to tell them which side won.
The investors who navigated 1999 recognized that a real technology and an overpriced market can coexist, and they sized their exposure so that the capital cycle resolving first would not wipe them out before the technology cycle resolved second. That is the framework both of these promo campaigns are implicitly arguing about, and neither one will say it out loud because the promo format needs you to pick a side.
July 29 is the first earnings window where both theses get tested against the same data, and neither data point will resolve the structural question, because the structural question is whether the debt service and the productivity gains arrive on the same schedule and that question takes years to answer.
See the guides index for more publisher and thesis explainers.