Investors should monitor SpaceX for potential entry points following its recent IPO, but remain cautious of bearish pressure later this year as insider lockup periods expire. Consider exposure to the "NeoCloud" theme by prioritizing companies like SpaceX that own physical AI infrastructure and data centers, which currently offer higher margins than software alone. Avoid heavy positions in Anthropic for now due to significant regulatory risks and government-mandated shutdowns of their primary models. Look for opportunities in the "Control Plane" sector, focusing on companies providing AI governance and auditability tools to help corporations navigate new national security regulations. Monitor OpenAI for a potential IPO filing, noting that their core inference business is now fundamentally profitable with $13 billion in annual revenue.

SpaceX (Private/Publicly Traded)

• SpaceX has recently undergone a massive strategic shift, evolving from a space exploration company into a major AI infrastructure and "NeoCloud" provider.
• The company built two massive supercomputer data centers, Colossus 1 and 2, which are being monetized by providing compute power to major AI labs like Anthropic and Google.
• Financial Performance: Following its IPO, the stock closed at $201.80, up 49% from its initial price. The company is now valued at $2.6 trillion, making it the 5th largest company in the world by market cap, surpassing Amazon.
• Acquisitions: SpaceX acquired the AI coding startup Cursor for $60 billion. Cursor is reportedly hitting a $4 billion revenue run rate and growing 7x year-over-year.
• New Models: SpaceX/Cursor is teasing a new frontier model (under the Composer brand) that is expected to be 10-20x more powerful than previous versions and competitive with GPT-5.5.

Takeaways

• Infrastructure as a Moat: SpaceX is successfully pivoting its capital-intensive hardware (data centers) into high-margin cloud revenue.
• Valuation Risks: Critics point out that SpaceX is valued higher than Amazon despite having significantly less revenue (roughly 1/40th). Investors should watch for "bearish pressure" later this year as employee and insider stock lockup periods expire.
• Trillionaire Status: Elon Musk’s wealth is heavily tied to his 46% stake in SpaceX; however, liquidity remains a challenge as selling large portions could tank the market price.

Anthropic (Private)

• Anthropic is currently embroiled in a major regulatory battle with the U.S. Department of Commerce.
• The government forced a shutdown of their latest models, Mythos and Fable 5, due to national security concerns and a "jailbreak" vulnerability that could allow the models to be used for cyberattacks.
• Geopolitical Friction: The administration is concerned that a South Korean telecom company with ties to the Chinese government gained access to Anthropic’s technology.
• Financial Plans: Anthropic has reportedly filed confidentially for an IPO later this year, though the current regulatory standoff may complicate the timeline.

Takeaways

• Regulatory Risk: This situation highlights a "new world order" where AI models are treated as national security assets. Anthropic’s failure to manage government relations is seen as a significant business risk.
• Product Reliability: The "switch-off" of their models by the government creates uncertainty for enterprise clients who require "business continuity."

OpenAI (Private)

• Leaked audited financials reveal a complex but growing business. In 2025, OpenAI reported $13 billion in revenue.
• While the company showed a massive net loss ($38.5 billion), much of this was a one-time non-cash accounting charge related to restructuring into a public benefit company.
• Profitability: Stripping away training and one-time costs, OpenAI appears to be making a profit on "inference" (the cost of running the AI for users), with $13 billion in revenue against $7.5 billion in direct costs.
• Cash Position: The company is well-capitalized with $73 billion in cash and marketable securities.

Takeaways

• IPO Outlook: Like Anthropic, OpenAI has filed confidentially for an IPO. However, with a massive cash pile, they may choose to stay private longer to avoid the volatility currently seen in the public AI markets.
• Efficiency: The data suggests that the core business of "selling tokens" is becoming fundamentally profitable, even if research and training costs remain high.

Investment Themes & Sectors

AI Infrastructure and "NeoClouds"

• There is a massive shift from just building models to owning the compute infrastructure. Companies like SpaceX are proving that owning the "iron" (data centers) is a faster path to revenue than software alone.

Agentic AI Economics

• The industry is moving from "token subsidy" (cheap AI) to "token scarcity" (expensive, high-reasoning AI).
• Investment Insight: Look for companies focusing on efficiency and routing—using the cheapest model possible for a specific task to preserve margins.

National Security & Export Controls

• AI is now a "controlled technology." The U.S. government is actively intervening to prevent foreign nationals (specifically from China) from accessing frontier models.
• Risk Factor: Any AI company with heavy international exposure or "open" distribution models faces increased "de-platforming" risk by federal regulators.

The "Control Plane"

• As identified by investor Chamath Palahapitiya, the next phase of value creation is the application layer.
• Takeaway: Investors should look for companies providing governance, auditability, and control—tools that help big corporations use AI safely and legally.