Ed Ludlow 3.0 82 ideas

Boeing's Space Launch System (SLS) moon rocket is way over budget and behind schedule, faces competition from private companies like Space X and Blue Origin, and the Trump administration is considering replacements, indicating an uncertain future for Boeing's space contracts and potential risks.

Broadcom announced a long-term agreement with Google to supply Tensor Processing Units (TPUs) and confirmed an expanded partnership with AI lab Anthropic, tripling capacity to 3.5 gigawatts. This deal is interpreted as a vote of confidence in Broadcom's architecture and evidence it can close the gap with Nvidia over the longer term. Anthropic's explosive growth (revenue run rate from $9B to $30B) directly drives demand for Broadcom's chips. Positive catalyst and affirmation of competitive positioning in the critical AI infrastructure supply chain, leading to pre-market stock gains of ~3.5%. Execution risk in scaling production to meet demand; potential supply chain disruptions (e.g., helium shortage).

The speaker explicitly names Lam, KLA, Applied Materials, and ASML as the companies that supply the machines required for chip fabrication. The speaker states that spending for the massive Terafab project "is going to be on others... on the chip equipment makers, not necessarily on Intel." If the Terafab project moves forward toward its extraordinarily ambitious scale (requiring trillions in spending), the companies that manufacture the core fabrication equipment would be the direct and primary capital goods beneficiaries. WATCH because this project represents a potential, albeit highly uncertain, mega-capacity expansion cycle for leading-edge logic fabrication, which would drive demand for this select group of critical equipment suppliers. The Terafab project is delayed, scaled down significantly, or canceled.

The speaker states Intel's story is that it has failed to scale its leading-edge processes or find customers other than itself, with no evidence of high-volume throughput for its next-gen nodes. The speaker questions whether providing planning expertise to the Terafab project will help Intel recover from its "incredible fall." Intel's participation in the Terafab project is framed as advisory (planning, structure) rather than as a primary capital expenditure recipient. The actual spending will flow to equipment makers, not Intel. This does not solve Intel's core operational and competitive deficiencies. AVOID because the announced partnership does not appear to directly address the fundamental, unattractive problems in Intel's manufacturing business. The company's recovery thesis remains unproven and is not clearly strengthened by this news. The Terafab project materializes at scale much faster than expected and Intel secures a lucrative, high-volume manufacturing role within it.

The speaker confirmed reports that Amazon is in acquisition talks with Globalstar, noting Globalstar's stock rallied ~9% on the news. He detailed that Globalstar's core value is its licensed spectrum and its established global regulatory authorizations and corresponding ground infrastructure. Acquiring these assets would provide Amazon with a critical "go-to-market" advantage for its own satellite internet service (Project Kuiper), bypassing lengthy regulatory processes and infrastructure build-out. This directly positions Amazon to compete more effectively with SpaceX's Starlink. WATCH because the deal is not finalized but represents a significant, high-stakes strategic move in the competitive satellite internet sector. Success would materially advance Amazon's space ambitions. The deal talks could fall through. The strategic value is contingent on regulatory approval of the transfer of spectrum licenses.

Microsoft is in exclusive talks with Chevron and Engine No. 1 on a long-term deal to underpin a giant power plant in West Texas, providing electricity for a large data center campus. Hyperscalers face a severe bottleneck in securing guaranteed energy supply for their announced AI data center builds. This deal is a direct attempt to lock in that critical input. WATCH as this exemplifies the intense, strategic scramble by leading tech firms to secure energy assets, which is becoming a primary constraint and competitive factor in AI infrastructure expansion. Deal talks fall through, or the scale of energy demand outpaces the ability of such bespoke deals to meet it, leaving capacity constraints unaddressed.

SpaceX confidentially filed for an IPO to fund its future business plan centered on space-based data centers and the Starship vehicle, which is critical for its economics (dollar per kilogram) and lunar ambitions. The company needs significant capital to develop and prove Starship's capabilities, which are foundational to its next-phase business model. An IPO provides that capital. WATCH because the filing indicates a major capital event and a pivotal attempt to fund an ambitious, unproven, but potentially transformative business plan. Starship fails to achieve its promised low-cost economics or faces further developmental delays, undermining the core rationale for the capital raise.

A Fortune report states Anthropic is testing a new AI model with significant offensive cybersecurity capabilities. The report caused Crowdstrike, Palo Alto, and Cloudflare to sell off sharply (down 3-6.5%). The model's advanced offensive capabilities are seen as a direct threat to the defensive value proposition of incumbent cybersecurity software companies. This introduces a new, near-term disruption risk to the cybersecurity sector, justifying a negative view. The model is not yet released, and its practical impact may be overstated. Defensive AI could also improve alongside offensive AI.

Fortune reported on an unreleased Anthropic model with offensive cybersecurity capabilities, and CrowdStrike and Palo Alto stocks are down as a result. The model's offensive prowess threatens the defensive business models of established cybersecurity companies, leading to negative market sentiment. This development introduces significant risk and makes these stocks unattractive in the near term, warranting avoidance. The report may be inaccurate, the model's capabilities could be overstated, or cyber companies might adapt to mitigate the threat.