A major M&A battle for Warner Bros. Discovery (WBD) is underway, with an all-cash takeover offer from Paramount at $30/share. NVIDIA (NVDA) has received approval to sell its H200 chips to China, a policy shift that could add $8 billion to $20 billion in new annual revenue. A compelling long-term investment theme is emerging around space data centers, with private company SpaceX seen as the key player to watch for a potential future IPO. For speculative public market exposure to this sector, investors are looking at companies like AST SpaceMobile (ASTS). Finally, monitor the upcoming AI IPO pipeline, which is expected to begin with enterprise-focused Cohere, as a key test of market appetite for new AI stocks.

Warner Bros. Discovery (WBD) & Paramount (PARA)

• There is a major M&A battle underway. Paramount has launched a hostile, all-cash takeover bid to acquire Warner Brothers for $77.9 billion.
• This is the latest in a series of offers from Paramount's David Ellison, who has progressively increased the price and the cash component of the deal. The offers have evolved from $19/share (60% cash) to $30/share (100% cash).
• The Warner Brothers board is considering the Paramount offer against a competing offer from Netflix (NFLX).
• The decision is not just about the highest price. The board must calculate the expected value of each bid, which involves discounting the offer price by the perceived risk that the deal won't actually close.
  • Key risks include financing risk (will the buyer have the money?) and regulatory risk (will the FTC or other bodies block the deal?).
• Paramount's financing for the bid is reportedly backed by a global consortium, including $11.8 billion from the Ellison family, $24 billion from sovereign wealth funds in Saudi Arabia, Qatar, and Abu Dhabi, and $1 billion from China's Tencent, among others.
• An underrated asset mentioned is Warner Brothers' "vault of masculine cinema," including franchises like Rambo, Mad Max, The Matrix, and numerous Clint Eastwood and Steven Seagal films.

Takeaways

• This is a classic M&A arbitrage situation. The stock price of WBD will be highly sensitive to news about the competing bids from Paramount and Netflix.
• Investors should monitor for any signs that one deal is more likely to get regulatory approval than the other, as this is a key factor for the board's decision.
• The increasing all-cash nature of Paramount's offer reduces the risk for WBD shareholders, as they wouldn't be left holding stock in the newly combined company, which could trade down.
• The involvement of numerous international funds, including from the Middle East and China, is now standard for large-scale M&A and is a key part of the modern deal-making playbook.

NVIDIA (NVDA)

• The Trump administration has announced it will allow NVIDIA to sell its H200 chips to approved customers in China. This is a reversal of previous, more restrictive policies.
• The U.S. government will reportedly receive a 25% cut of these sales, an unusual export duty.
• The H200 is a generation behind NVIDIA's top-tier Blackwell and upcoming Rubin chips, but it is more powerful than the previously nerfed chips allowed for sale in China.
• The market reaction was muted, with NVDA stock being down slightly on the day of the news but up nearly 2% in after-hours trading. The podcast suggests this news may have already been priced in.
• NVIDIA's CFO, Collette Kress, stated that if geopolitical issues subside, the company could ship between $2 billion and $5 billion worth of chips to China per quarter.
• There's a strategic debate around this decision:
  • Bull Case: Selling chips to China gets their developers "addicted to the American technology stack" (CUDA ecosystem), floods their market with cheaper, better chips, and makes it economically difficult for them to prop up domestic competitors like Huawei. The revenue also funds further US R&D.
  • Bear Case: This move could accelerate China's AI capabilities. The hosts point to the example of Tesla (TSLA) in China, where local competitors like BYD and Huawei were able to learn and eventually leapfrog Tesla's technology.

Takeaways

• This policy change is a significant potential revenue driver for NVIDIA, potentially adding $8 billion to $20 billion in annual sales from China alone.
• Despite the 25% government cut, NVIDIA's massive gross margins (mentioned as 73.4% on $57 billion in sales in the most recent quarter) mean they can easily afford this and still be highly profitable.
• Investors should watch for China's actual purchasing volume. The Chinese government could still push back and favor domestic champions, limiting the upside for NVIDIA.
• The long-term risk is whether this policy helps or hurts the U.S. in the AI race. If China uses this access to accelerate its own domestic chip industry, it could create a formidable competitor for NVIDIA down the line.

Space Data Centers (Investment Theme)

• Prominent investor Gavin Baker described data centers in space as "the most important thing that's going to happen in the world in the next three to four years."
• The Bull Case (First Principles):
  • Power: Satellites in a sun-synchronous orbit have access to constant, intense solar power (6x more irradiance than on Earth) without the need for expensive batteries.
  • Cooling: Space is near absolute zero. Cooling, a massive cost and weight component of terrestrial data centers, is effectively "free" by using a radiator on the dark side of the satellite.
  • Networking: Laser links between satellites in a vacuum are faster than fiber optic cables on Earth.
• The Bear Case (Practical Challenges):
  • Heat dissipation in a vacuum is a complex physics and material science problem, not as simple as just "putting a radiator on the dark side."
  • Even in sun-synchronous orbits, satellites can occasionally pass through Earth's shadow, potentially requiring some battery backup.
  • The cost and complexity of launching, maintaining, and upgrading hardware in space are immense.
• SpaceX is seen as the company best positioned to make this work, using a system derived from its Starlink satellites. One Starship launch could potentially deploy 30 megawatts of inference capacity.
• AST SpaceMobile (ASTS) was mentioned as being up 27% in the last five days, suggesting growing investor excitement in the space-related technology sector.

Takeaways

• This is a nascent but potentially massive investment theme. While the timeline is debated (3-4 years vs. 10-20 years), the core idea of moving power-hungry compute to a location with free power and cooling is compelling.
• SpaceX is the key private company to watch in this space. An IPO or more public information about their data center plans could be a major catalyst. The podcast notes that a space data center narrative could significantly boost its valuation ahead of a potential 2026 IPO.
• Investors interested in this theme can look at publicly traded space companies like ASTS, but should be aware that this is a high-risk, long-term play.

AI Foundation Models & Companies

• Market Sentiment: There's a growing debate about the long-term defensibility of foundation models.
  • Mark Benioff (Salesforce CEO) argues that LLMs are becoming a commodity, like "new disk drives" that can be hot-swapped for the cheapest and best provider.
  • Michael Burry (of "The Big Short" fame) is extremely bearish on OpenAI, calling it the "next Netscape" and "hemorrhaging cash." The podcast hosts push back, noting OpenAI's revenue is vastly larger than Netscape's ever was.
• IPO Pipeline: A rumored IPO sequence for the major AI labs is:
  1. Cohere: Focused on the enterprise market, which could be a unique and compelling narrative for public markets.
  2. Anthropic
  3. OpenAI
• Custom Hardware: OpenAI is reportedly working with Broadcom (AVGO) to build its own custom AI chip. The goal is a chip that is more flexible than Google's TPUs to accommodate rapid changes in AI algorithms (e.g., moving from dense to mixture-of-experts models).
• Private AI Companies Mentioned:
  • Unconventional AI: A new startup from serial entrepreneur Naveen Rao that raised a massive $500 million seed round. They aim to build a new type of computer, 1000x more efficient for AI, by moving away from the digital "bits" abstraction.
  • Fal.ai: Raised a $140 million Series D. They provide a platform for running various AI models and have seen 8x revenue growth, driven surprisingly by image editing even more than AI video.

Takeaways

• The "model is the moat" thesis is being challenged. Investors should consider that the value in AI may shift from the foundation model creators to the companies that apply these models or provide the underlying infrastructure.
• Keep an eye on the upcoming AI IPOs, starting with Cohere. Its performance will be a key test of public market appetite for AI companies that are not OpenAI.
• The move by OpenAI to build custom chips highlights a key trend: the biggest AI players are vertically integrating to control their own hardware destiny and reduce dependence on NVIDIA. This is a long-term risk for NVIDIA's dominance.
• The massive funding rounds for new hardware startups like Unconventional AI show that venture capital believes there is still room for fundamental disruption in the AI chip space beyond the current players.

Boom Supersonic (Private)

• Boom, known for developing a supersonic passenger jet (Overture), has announced a new product line called SuperPower.
• SuperPower consists of 42-megawatt natural gas turbines designed to power AI data centers. This is an "aeroderivative" turbine, meaning it's based on the same core jet engine technology being developed for their supersonic plane.
• The company has raised a $300 million funding round, which they state is the last equity round they will ever need to raise.
• Crusoe Energy is the launch customer for the SuperPower turbines.
• This is a strategic pivot to generate revenue and de-risk the much more capital-intensive development of the Overture jet. The power turbine business is seen as "easy mode" compared to building a passenger aircraft.

Takeaways

• This is a brilliant strategic move that significantly increases Boom's chances of success. By entering the high-demand market for data center power, they create a cash flow stream to fund their ultimate vision of supersonic travel.
• For private investors, this makes Boom a much more attractive and less risky proposition. The company is no longer a binary bet on supersonic flight; it now has a tangible, high-growth business in the AI infrastructure space.
• This highlights a broader trend: the immense energy needs of AI are creating new markets and opportunities for companies in the energy and industrial sectors, not just tech.

Brex (Private)

• Brex announced a massive $5.6 billion commercial card partnership with Fifth Third Bank (FITB).
• This partnership will allow Fifth Third's huge base of commercial customers to get a Brex-branded corporate card and use Brex's software and AI-powered expense management platform.
• A key competitive advantage for Brex is that they built their own financial infrastructure from the ground up, rather than using third-party providers like Stripe Issuing or Marqeta.
• This "full-stack" approach gives them the flexibility to forge complex partnerships like this one and to operate globally in over 200 countries with local currency issuance.

Takeaways

• This partnership is a massive distribution channel for Brex, giving them access to a huge portion of the U.S. business market that still uses legacy banking products.
• Brex's strategy of investing heavily in its own core infrastructure is paying off, creating a moat that is difficult for competitors to replicate.
• For traditional banks like Fifth Third, this is a smart way to modernize their offerings and compete with fintechs by partnering with them instead of trying to build everything in-house. This could be a model for other regional banks.