In June 2025, James Altucher released a briefing called “The AI Mothership Has Landed.” The thesis was simple and enormous. Elon Musk was building a supercomputer in Memphis, Tennessee, called Project Colossus. Altucher argued it would usher in what he called Artificial Superintelligence — a second wave of AI that would make ChatGPT look like a calculator.

The briefing ran through GlobeNewswire on June 4, 2025. A follow-up presentation appeared in August. The claims were specific. Musk had surpassed OpenAI, Meta, and Microsoft with a covert project few had heard of. Jensen Huang had acknowledged the facility. Trump had repealed Biden’s Executive Order 14110 on his first day in office, clearing the regulatory path. A $500 billion AI infrastructure plan called Stargate was announced. Altucher predicted Musk would unveil a 10x power upgrade to Colossus within weeks.

The numbers were the part that got attention. Altucher cited estimates that artificial intelligence would create $20 trillion in new wealth per year by 2030. He quoted scientists projecting $14 quadrillion in cumulative windfalls over 20 years. The headline prediction was a trillion-dollar call on Musk himself. “It’s not on a single company,” Altucher said. “But on a single man — Elon Musk.”

What Project Colossus Actually Is

The Memphis facility is real. Musk’s xAI division built a supercomputer cluster using Nvidia GPUs at a site in Memphis, scaling to what Nvidia CEO Jensen Huang publicly confirmed as one of the largest AI training clusters in the world. The project involved gigawatt-scale power consumption — enough electricity to power a small city — dedicated to training large language models.

Altucher’s framing connected three threads. First, the hardware: the GPU cluster itself, which he described as a manufacturing facility for intelligence. Second, the deregulation: Trump’s repeal of Biden’s AI executive order, which had imposed safety reporting requirements on large AI models. Third, the geopolitical angle: Altucher cited Putin’s warning that whoever leads in AI would rule the world.

The briefing positioned Altucher as someone who had been inside the AI world since the 1990s. He worked on a chess program at Carnegie Mellon that became IBM’s Deep Blue. He ran algorithmic trading funds. He published an academic paper on AI at Cornell before he could legally drink. The AI expert profile and career arc behind that claim is the reason the presentation lands: this is not a commentator reading press releases but someone who built the predecessor system. The AI Mothership presentation leaned on that biography to frame the Colossus analysis as the perspective of someone who understood the technology from the inside.

The Musk Pattern

The AI Mothership presentation followed a pattern Altucher had used before. Find a technology adoption curve before the market prices it in. Identify the central figure. Attach a number that sounds absurd at the time. Wait.

The Facebook call in 2007 followed this structure. Altucher said Facebook would be worth $100 billion when the company was valued at under $1 billion. The Apple call in 2010 followed it: he predicted Apple would be the first $1 trillion company when it was at $320. The Bitcoin call in 2013 followed it: he sold his book for 0.1 BTC when the price was $114. The track record of dated big calls documents all three with verifiable primary sources, and the structure is what matters here.

The AI Mothership applied the same structure to Musk’s infrastructure. The $20 trillion figure was the equivalent of the $100 billion Facebook claim — a number large enough to seem reckless at the time, anchored to a specific year and a specific adoption curve. The Musk bet was the equivalent of naming the company: Altucher was saying the individual mattered more than the technology category.

Where the Thesis Stands

The AI Mothership ran as a free presentation that introduced readers to Altucher’s paid products at Paradigm Press. The August 2025 follow-up pushed the same thesis with added urgency. By 2026, the infrastructure Altucher described had continued to expand. SpaceX acquired xAI in early 2026, bringing the Colossus data center, the Grok language model, and the X social network under one corporate roof. The SpaceX IPO on June 12, 2026, gave public market investors a way to own a piece of the AI infrastructure stack Altucher had described a year earlier.

The $20 trillion annual wealth creation figure remains a projection, not a measurement. The $14 quadrillion number is a theoretical upper bound cited by optimistic analysts, not a consensus estimate. What is verifiable is the infrastructure buildout: the Memphis facility expanded, the GPU count grew, and the power consumption scaled as Altucher predicted it would.

The presentation matters now because it was the template. Deep Blue 2.0, launched July 10, 2026, uses the same architecture: Altucher’s AI credentials as the credibility engine, a specific technology as the thesis, and a stock-picking tool as the product. The AI Mothership was the story. Deep Blue 2.0 is the tool that operationalizes the story.

Read the June 2025 briefing and the July 2026 Deep Blue 2.0 landing page side by side, and the throughline is visible. Both start with Altucher’s Carnegie Mellon background. Both cite the Deep Blue lineage. Both argue that the person who understood AI before it was commercial is the person worth listening to now that it is. The Choose Yourself philosophy is the brand that ties the two together: the man behind that philosophy is the same person who worked on Deep Blue in the 1990s and built a stock screener around it in 2026. The AI Mothership told you why AI mattered. Deep Blue 2.0 tells you which stocks to buy because of it.

That is the progression: the ideas are the same, and the product evolved to operationalize them. More guru files are collected in the Guru Files.