June 8, 2026 Meeting

Introduction

We held the latest meeting of The Investing Group on Monday, June 8, 2026 at the Union League Club of Chicago in Room 816.

The meeting continued the open discussion format that has become the defining characteristic of the group. Conversation moved naturally across market structure, AI infrastructure, semiconductors, space, robotics, healthcare, crypto, real estate, energy, and several emerging speculative themes.

Several recurring themes connected nearly every topic discussed throughout the evening. The first was that infrastructure matters more than applications. The second was that scarcity continues to drive value, whether the scarce asset is compute, power, land, semiconductors, data, distribution, or public float. The third was that markets continue moving toward greater speculation, faster information flow, and increasingly reflexive behavior.

1. Chinese Micro-Cap Mania

The meeting opened with one of the most extreme examples of speculative market behavior discussed in any recent session.

Attention centered on Inno Holdings (INHD), a Hong Kong-linked electronics and used-phone micro-cap company that surged approximately 3,660% in a single trading session on June 8, rising from a prior close of roughly $1.05 to close at $39.49. During the session, the stock traded above $60 and experienced repeated volatility halts, illustrating the extraordinary degree of speculation and price dislocation that can emerge in low-float securities.

The catalyst for the move was a recently announced $3 million Development Services Agreement with an undisclosed Hong Kong-based AI provider to build an AI sales-agent system for the company's used-phone business. The announcement centered on the adoption of AI tools within the company's existing business, yet it sparked one of the most dramatic speculative moves seen in the market this year.

The conversation focused on how these moves increasingly resemble financial engineering exercises rather than business developments. Many micro-cap companies continue using U.S. exchanges as capital-raising vehicles, often combining low floats, promotional narratives, reverse splits, at-the-market equity programs, and speculative themes to generate extraordinary price volatility.

INHD itself provided a useful case study. The company had completed two reverse splits within the previous six months to maintain Nasdaq listing compliance and recently announced a sizable at-the-market equity facility. Sudden increases in market capitalization can dramatically expand future financing flexibility, creating incentives that may not always align with long-term business fundamentals.

Nasdaq subsequently halted trading after the close under a T12 code pending the release of additional information. With the stock frozen at the halt price, short sellers remained exposed to ongoing borrow costs while having limited ability to actively manage positions until trading resumed. The halt served as another example of how liquidity constraints, float dynamics, and market mechanics can sometimes become as important as the underlying business itself.

The group also discussed how the recent elimination of the Pattern Day Trader designation and associated $25,000 minimum equity requirement may be contributing to increased speculative activity. While it remains far too early to establish a direct connection, easier access to active trading could further amplify volatility in already thinly traded securities.

A broader takeaway emerged that market structure itself is becoming a larger driver of price action. In certain corners of the market, float dynamics, liquidity constraints, retail participation, short positioning, borrow costs, and trader behavior appear capable of overwhelming traditional fundamental analysis for extended periods of time. The INHD episode served as a vivid illustration of how market mechanics can sometimes become the primary story.

2. Iran, Oil Markets, and Demand Destruction

The group revisited the conflict involving the United States, Israel, and Iran and its implications for global energy markets.

A significant portion of the discussion focused on the sharp reaction in crude oil following the escalation of hostilities and the disruption of shipping through the Strait of Hormuz. Oil prices briefly moved into triple-digit territory as investors weighed the possibility of prolonged supply disruptions across one of the world's most important energy corridors.

The conversation also examined the performance of broader financial markets during the conflict. Discussion focused on investor expectations, economic growth, corporate earnings, and the factors shaping market behavior during periods of geopolitical uncertainty.

The group examined the concept of demand destruction and the long-term outlook for global oil consumption. Rising electric vehicle adoption, continued improvements in fuel efficiency, expanding renewable energy capacity, and broader technological changes were cited as factors

that could place pressure on future demand growth over time. The discussion explored how these trends may influence the long-term trajectory of energy markets and the outlook for future oil demand.

Discussion also touched on how higher energy prices can create benefits for certain segments of the U.S. energy industry by increasing the value of domestic production, while additional production from major exporting nations may help moderate prices as supply conditions evolve.

The conversation concluded with observations about the importance of critical energy infrastructure, global supply routes, geopolitical risk, and the wide range of factors that continue to shape energy markets and investor expectations.

3. Anthropic, Recursive Self-Improvement, and the Future of AI

A significant portion of the evening focused on Anthropic’s recent paper, "When AI Builds Itself" (https://www.anthropic.com/institute/recursive-self-improvement), and what it may imply for the future trajectory of artificial intelligence.

The paper examines whether AI systems may be approaching a stage where they can contribute to improving future generations of AI models, potentially accelerating development cycles beyond traditional human-led progress. Anthropic noted that recursive self-improvement is not inevitable and that current systems have not yet reached that threshold, but presented evidence suggesting that AI-assisted development is already playing a meaningful role within leading AI labs.

Particular attention was paid to Anthropic’s internal data showing that AI systems now contribute a substantial share of the code used to build and improve future models, alongside major gains in engineering productivity. The findings were viewed by some attendees as broadly consistent with themes previously outlined by Leopold Aschenbrenner regarding AI scaling and the possibility of faster development cycles in the years ahead. At the same time, Anthropic framed the paper primarily through the lens of safety, governance, and the potential need for industry coordination if self-improving systems become a reality.

The conversation frequently returned to the relationship between AI progress, semiconductor demand, and energy infrastructure. Members noted that even modest advances in AI capabilities could translate into significantly greater demand for compute resources, semiconductors, data centers, and power generation. While the paper itself focused on potential recursive self-improvement scenarios, many viewed the infrastructure implications as one of the most important investment takeaways.

The group also explored how more capable AI systems may affect software development, cybersecurity, scientific research, and economic productivity. While opinions differed on the pace of adoption, there was broad agreement that ownership of productive assets, infrastructure, and intellectual property could become more important if AI automates a larger share of knowledge work.

Additional attention was given to Anthropic’s restricted Mythos model and the cybersecurity implications of advanced AI systems. Members examined how these tools could enhance software development and research while also expanding the ability to identify vulnerabilities, highlighting both the opportunities and challenges associated with continued advances in AI capability.

4. Compute Scarcity, Jevons Paradox, and AI Infrastructure

Compute scarcity remained one of the dominant themes of the evening.

The discussion explored recent innovations that significantly reduce memory requirements and improve model efficiency. Some viewed these advances through the lens of Jevons Paradox, the economic observation that greater efficiency often leads to higher overall consumption rather than lower demand.

Declining token costs, improved model performance, and more efficient architectures were cited as examples. While the cost of generating AI outputs has fallen dramatically in recent years, overall AI spending and usage have continued to rise as lower costs enable larger workloads, broader adoption, and new categories of applications. Attention was also given to the growing role of agentic systems, which can consume substantially more compute than traditional chatbot interactions.

The conversation expanded into power generation, GPU manufacturing, data center construction, networking infrastructure, semiconductor production, and the rapid buildout of AI capacity across the industry. Recent advances in memory efficiency and model optimization were viewed as important developments, but many felt they were accelerating demand for infrastructure rather than reducing it.

Particular attention was given to power as one of the most important constraints facing the industry. Discussion focused on grid capacity, transmission infrastructure, transformer shortages, and the growing interest in dedicated power solutions ranging from natural gas generation to nuclear energy. Several large technology companies have already announced major investments and partnerships aimed at securing long-term power supplies for future AI deployments.

A recurring theme throughout the discussion was that AI remains constrained less by software than by infrastructure. Compute capacity, power generation, networking, cooling systems, semiconductor manufacturing, and grid connectivity continue to play a central role in determining how quickly AI capabilities can scale.

5. NASA Moon Base Missions and Orbital Infrastructure

The group reviewed NASA's recent Moon Base announcement, which outlined plans to establish a sustained long-term presence on the lunar surface. The announcement offered a more detailed look at NASA's roadmap for lunar infrastructure development and the broader effort to expand human activity beyond Earth.

Discussion focused on NASA's phased roadmap, which begins with robotic missions, cargo deliveries, and surface operations before progressing toward a permanent base near the Moon's south pole. Particular attention was given to the strategic importance of the region due to its large deposits of water ice, which could support future habitation, fuel production, and resource utilization. NASA has already committed roughly $1 billion in initial contracts across several commercial partners, including Blue Origin's Blue Moon cargo lander program, Astrobotic's Griffin lunar lander, and lunar terrain vehicle contracts awarded to AstroLab and Lunar Outpost. NASA's first Moon Base mission is currently targeted for no earlier than fall 2026, with the broader program extending through the early 2030s. The surface base itself is expected to carry an estimated cost of roughly $20 billion.

Discussion connected these developments to the larger evolution of space infrastructure and the gradual expansion of commercial and industrial activity beyond Earth. The conversation covered satellite manufacturing, lunar launch and lander systems, orbital logistics, and the broader challenge of building self-sustaining infrastructure away from Earth. Additional context surrounding the Moon Base program highlighted the importance of long-term energy generation and resource utilization as key components of any permanent lunar presence.

The group viewed these efforts as early building blocks for a broader space economy. As with many large-scale infrastructure projects, attention was given to how foundational systems often require years of investment and development before their economic significance becomes fully apparent.

The broader takeaway was that lunar infrastructure has moved beyond a purely conceptual stage, with funding commitments, defined hardware programs, and mission timelines now in place. Execution remains the primary challenge, particularly given the large number of launches and coordinated missions required over the coming decade, but the program was viewed as a meaningful step toward the long-term development of space-based infrastructure and industry.

6. SpaceX, the Commercial Space Ecosystem, and the IPO Opportunity

SpaceX dominated a significant portion of the meeting.

The conversation centered on valuation, float dynamics, index inclusion, retail participation, and long-term business prospects ahead of the company's anticipated IPO. A key reference point was recent valuation work by NYU finance professor Aswath Damodaran, who is well known for his valuation analysis and investing blog. Damodaran had recently published a series of analyses valuing SpaceX at approximately $1.2 trillion to $1.3 trillion, providing a framework for much of the discussion. The conversation also examined the possibility of an "Elon premium" above traditional valuation models, reflecting the market's willingness to assign higher multiples to companies associated with Elon Musk's track record of building category-defining businesses. Reference was made to Tesla (TSLA), where Damodaran's valuation work often proved more conservative than the market's eventual assessment of the company's long-term growth prospects. The group also reviewed Polymarket prediction markets tied to SpaceX's anticipated public debut, which assigned varying probabilities to valuation tiers ranging from roughly $2.0 trillion to $3.0 trillion. The $2.4 trillion level emerged as one of several frequently discussed scenarios for the company's valuation shortly after public trading began.

Another theme was the extent to which SpaceX now sits at the intersection of aerospace, communications, and AI infrastructure. Discussion referenced the company's merger with xAI and major compute agreements involving Anthropic and Alphabet (GOOGL). The conversation also highlighted SpaceX's Colossus data center infrastructure in Memphis, which had emerged as one of the largest AI computing facilities in the world. The broader view was that SpaceX's long-term opportunity may extend beyond launch services, satellites, and communications into the rapidly expanding market for AI compute and infrastructure.

Discussion also focused on the company's expected public float structure. With only a limited percentage of shares anticipated to trade publicly and lockup restrictions potentially constraining additional supply, many viewed the offering as having the potential to become one of the most volatile large-cap IPOs ever brought to market. The conversation repeatedly returned to a simple supply-and-demand framework: substantial investor interest combined with a limited supply of available shares could create significant price volatility.

The discussion also expanded beyond SpaceX itself and into companies that may benefit indirectly from its success. Rocket Lab (RKLB) received attention as a potential sympathy trade. While considerably smaller than SpaceX, the company operates within many of the same commercial space markets and could benefit from renewed investor interest in the sector. The speculative nature of many publicly traded space companies was also discussed. Rocket Lab's

long-term valuation remains heavily dependent on the successful development and commercialization of its medium-lift Neutron rocket, which had not yet entered operational service and would likely face the normal testing, development, and execution risks associated with new launch systems. Planet Labs (PL) was also discussed because of its satellite imaging capabilities and growing role in space-based data collection, though the company remains relatively early in its commercial development. More broadly, the conversation noted that many publicly traded space companies continue to trade at elevated revenue multiples that assume substantial future growth and execution, leaving the sector highly speculative despite the attractive long-term opportunities.

A recurring theme was the value of second-order thinking. Rather than focusing exclusively on SpaceX, the conversation examined how suppliers, partners, infrastructure providers, and adjacent businesses may also benefit as commercial space activity continues to expand. The broader takeaway was that major technological shifts often create opportunities throughout an ecosystem, not solely within the headline company driving the narrative.

7. Drone Companies, Defense Technology, and Government Investment

The group also discussed reports that the U.S. government is exploring a mix of debt and equity financing for select domestic drone companies as part of a broader effort to expand production capacity and reduce reliance on Chinese-made components.

The conversation touched on companies across the sector, including AeroVironment (AVAV), Unusual Machines (UMAC), Red Cat Holdings (RCAT), Kratos Defense & Security Solutions (KTOS), and Ondas Holdings (ONDS). Discussion also covered the sharp market reaction following the reports, with several drone-related stocks posting double-digit gains as investors speculated on which companies could ultimately receive government support.

The topic highlighted the growing convergence of defense, AI, autonomy, robotics, aerospace, and advanced communications systems. Modern military capabilities now rely heavily on software, autonomous platforms, sensor networks, communications infrastructure, and AI- assisted decision making.

The group also referenced Anduril as an example of a venture-backed defense technology company that has emerged as a significant competitor alongside traditional defense contractors. More broadly, the conversation noted how the U.S. government has already used direct investment and financing programs in strategic industries such as rare earth materials and semiconductor manufacturing, raising the possibility that similar approaches could become more common within defense technology.

The broader takeaway was that defense technology is becoming a much more software and AI- driven industry than in previous decades, with capital, innovation, and government support flowing toward autonomous systems and next-generation military technologies.

8. Robinhood AI Agents and the Automation of Retail Investing

Robinhood (HOOD) generated substantial discussion following its launch of Agentic Trading and a companion Agentic Credit Card, which allow AI agents to execute trades and make purchases on behalf of users. The group examined whether AI-assisted investing will improve outcomes for retail investors, how it fits within the evolution of algorithmic investing, and the risks associated with delegating financial decisions to systems that many users may not fully understand.

The structure of the rollout drew interest. Robinhood is providing the infrastructure that connects third-party AI agents to customer accounts through its platform, leaving the investment models themselves to outside developers. To limit risk, agent activity takes place through separate funded accounts isolated from a user's primary portfolio, users receive notifications of trades, agents preview orders before execution, and connections can be terminated instantly. Robinhood's disclosures also make clear that responsibility for investment decisions remains with the user rather than the platform.

The conversation highlighted how financial services are becoming more software-driven. Robinhood's use of Model Context Protocol infrastructure was viewed as another example of platforms positioning themselves as the operating layer through which AI systems interact with financial markets. The initial rollout focused on stocks and ETFs, with additional asset classes expected over time.

The discussion also focused on how rapidly markets are becoming automated. As AI tools assume a larger role in investment decisions, questions remain around transparency, accountability, risk management, and the extent to which investors should rely on automated systems. The group also discussed whether wider adoption of AI-assisted trading could elevate the importance of infrastructure, execution quality, and data advantages across financial markets.

Investing continues moving toward a technology-platform model, with software, automation, and algorithms playing a larger role in how capital is allocated and investment decisions are made.

9. Meta, Subscriptions, and Alternative Revenue Models

Meta Platforms (META) was discussed following the company's expansion of subscription offerings across Instagram, Facebook, and WhatsApp, including the recent rollout of its Meta One subscription platform and premium AI-related tiers.

The conversation explored whether consumers will ultimately pay for social media and AI- enhanced subscription services at meaningful scale or whether advertising will continue to dominate Meta's business model. While subscription products may create additional revenue opportunities, Meta's advertising business remains the company's primary economic engine, generating the overwhelming majority of total revenue.

The discussion also touched on the broader push among technology companies to develop revenue streams beyond advertising while simultaneously committing substantial capital toward AI infrastructure, data centers, and next-generation computing capacity. Meta's planned AI spending was viewed as another example of the industry's accelerating investment in artificial intelligence.

Meta's global network of users, extensive data assets, and ownership of some of the world's largest consumer platforms were viewed as significant advantages as AI capabilities continue advancing. The company's ability to deploy new products across billions of users gives it a scale few competitors can match.

The conversation concluded that large technology platforms with extensive user networks, proprietary data, and direct customer relationships may be well positioned to benefit from the continued development and commercialization of AI technologies, even if the long-term contribution from subscription products remains uncertain.

10. Nebius, Cerebras, Marvell, and AI Infrastructure Winners

A large portion of the lightning round focused on potential AI infrastructure beneficiaries.

Nebius Group (NBIS) generated significant interest following the late-May disclosure that it had become the largest equity position in Leopold Aschenbrenner's Situational Awareness fund. The position, which represented a 5.6% ownership stake in Nebius, was viewed as a strong vote of confidence in the company's role as a provider of AI cloud and compute infrastructure. The discussion also touched on Nebius' rapid growth, with revenue rising more than sixfold year over year as demand for AI capacity continues to expand.

Cerebras Systems (CBRS) was discussed following its recent public debut and strong entrance into the public markets. The company has attracted interest through its wafer-scale chip

architecture and AI systems designed as an alternative to traditional GPU-based approaches. Its growing adoption across AI training and inference workloads was viewed as another example of how the industry continues to support multiple hardware architectures beyond the dominant incumbents.

Marvell Technology (MRVL) was highlighted as a key infrastructure provider through its work in silicon photonics, custom silicon, networking, and connectivity solutions. As AI clusters continue growing in size and complexity, the ability to move data efficiently between chips, servers, and data centers remains a critical part of the technology stack.

The common theme across all three companies was infrastructure. Rather than competing directly at the model layer, they occupy important positions within the AI supply chain and benefit from the ongoing buildout of compute capacity. The conversation reinforced a recurring theme from previous meetings: some of the largest AI beneficiaries may ultimately be the companies providing the hardware, networking, and cloud infrastructure that enable the entire ecosystem to function.

11. Bitcoin, Crypto Treasuries, and Monetary Systems

Bitcoin returned as a topic after its decline into the low $60,000s, a sharp pullback from the record highs reached during the previous cycle.

The conversation examined the outlook for Bitcoin, Strategy (MSTR), and a growing group of crypto treasury companies that have adopted variations of the Bitcoin accumulation model.

One area of focus involved debt structures and future refinancing risks. Strategy's use of convertible debt served as a key example. While the company has successfully financed large Bitcoin purchases through capital markets, questions centered on how these structures may perform if Bitcoin prices remain weak for an extended period or if future refinancing occurs under less favorable conditions. The discussion also touched on the importance of maintaining a premium valuation to net asset value, which has historically supported the model's ability to raise additional capital.

Recent developments added another layer to the conversation. Strategy's willingness to consider limited Bitcoin sales marked a notable shift from its long-standing position of never selling its holdings, highlighting how treasury strategies may evolve as balance sheets and capital structures become more complex.

The group also revisited Bitcoin's original purpose as an alternative monetary system and whether growing institutional ownership has altered its role within financial markets. Some viewed Bitcoin as behaving more like a risk asset than it did in earlier years, while others

continued to see it as one of the most compelling long-term stores of value available to investors.

The conversation concluded with the observation that Bitcoin continues to occupy a unique position between speculative asset, monetary alternative, and emerging financial infrastructure. As public companies, institutional investors, and treasury vehicles expand their involvement, the debate over which of those roles will ultimately define Bitcoin remains unresolved.

12. The Chicago Bears, Hammond, Indiana, and Local Development

The group discussed the Chicago Bears' stadium plans following the team's recent board vote to advance development efforts in Hammond, Indiana. The decision marked the first formal board vote on any proposed stadium site and came after Illinois lawmakers failed to approve a stadium financing package, while Indiana moved forward with legislation that could support the project through a combination of development district and tourism-related tax revenues. Reports that Indiana could provide incentives approaching $1 billion, alongside roughly $2 billion of planned investment from the Bears themselves, helped frame the scale of the opportunity being considered.

The conversation examined infrastructure requirements, transportation challenges, political considerations, and real estate implications associated with a project of that magnitude. Questions centered on whether Hammond currently possesses the roads, transit connections, surrounding commercial development, and broader infrastructure necessary to support an NFL stadium and the activity that would accompany it. Some expressed skepticism that the area could accommodate a project of this scale without substantial additional investment and long- term planning.

Discussion also returned to Arlington Heights, where the Bears own approximately 326 acres acquired in 2021. While Hammond now appears to be the organization's primary focus, Arlington Heights remained relevant given the team's existing land ownership, prior planning efforts, and the significant resources already committed to the site. The contrast between the two locations highlighted the tradeoffs between established development plans and the incentives being offered to attract major projects.

The conversation reinforced how large-scale real estate developments are often shaped by infrastructure, public incentives, regulatory considerations, and political realities alongside the underlying economics of the project. The Bears' stadium search served as a useful example of how location decisions for major developments can evolve as financing opportunities and government support change over time.

13. Data Centers, Land Scarcity, and Community Pushback

The Virginia data center land-buyout story generated one of the more interesting conversations of the evening.

The group examined a situation in Ashburn, Virginia, where homeowners in a 143-home subdivision have explored negotiating collectively with data center developers in an effort to maximize the value of their properties. Reports surrounding the proposal have suggested a potential transaction exceeding $500 million, implying average proceeds of roughly $4 million per home, about four to five times higher than typical residential values in the neighborhood. The story highlighted how the rapid expansion of AI infrastructure can dramatically alter land values in areas located near major data center hubs.

The conversation highlighted the tension emerging between AI infrastructure development and local community concerns. While data centers bring substantial tax revenue and economic activity, nearby residents often raise concerns related to noise, power usage, water consumption, visual impact, and changes to neighborhood character. The Virginia example also underscored the practical challenges involved, including zoning approvals and the difficulty of securing agreement among large groups of property owners.

The discussion expanded to the physical constraints shaping the next phase of AI infrastructure buildout. Land availability, power generation, grid capacity, and permitting were viewed as important considerations as facilities continue to grow in scale. The group noted that some proposed developments around the country have faced significant opposition, with community organizations helping delay or block billions of dollars of planned projects.

The conversation also touched on potential long-term solutions, including dedicated power generation, remote data center construction near stranded energy resources, local hosting, and orbital computing infrastructure. While concepts such as space-based data centers remain early-stage, they reflect the search for alternatives as demand for compute, power, and suitable development sites continues to expand.

A recurring theme throughout the discussion was that AI infrastructure now resembles traditional industrial development in many respects. Access to land, power, permitting, and community support may prove just as important as advances in chips and software when determining where future capacity can be built.

14. Healthcare, Robotics, and the Next Wave

The meeting concluded with discussion around healthcare, robotics, and future innovation themes.

Intuitive Surgical (ISRG) was highlighted as one of the strongest examples of a company possessing a durable data moat. The value of the vast amount of procedural data generated through the company's robotic surgery platform was noted as a potential advantage as AI becomes more integrated into healthcare. For context, Intuitive has deployed more than 12,000 robotic surgical systems worldwide and has facilitated over 20 million cumulative procedures, providing a scale of real-world surgical data that few competitors can match. The company's continued investment in next-generation robotic systems and software capabilities further reinforced the view that data, workflow integration, and clinical experience may prove as important as the hardware itself.

The discussion also touched on Eli Lilly (LLY), medical device companies, healthcare optimization trends, and future drug discovery. Interest centered on how AI may accelerate pharmaceutical research and development, improve operational efficiency across healthcare systems, and support more personalized approaches to treatment. Eli Lilly's growing investment in AI-enabled drug discovery and research partnerships was cited as one example of how large healthcare companies are beginning to incorporate AI into core business operations.

Robotics generated significant interest as well. The conversation focused on the idea that AI's long-term impact may extend beyond software into physical-world systems through industrial robots, humanoid robots, logistics networks, autonomous machines, and broader automation platforms. Recent advances in model capabilities, falling hardware costs, and early commercial deployments were viewed as signs that physical AI is beginning to move from concept toward real-world implementation.

The group also noted that many of the most closely watched humanoid robotics companies remain private, leaving public market investors with more indirect exposure through infrastructure providers, semiconductor companies, component suppliers, and automation- focused businesses. The dynamic was compared to other emerging technology sectors where much of the value creation initially occurs in private markets before broader public participation becomes available.

The takeaway was that robotics remains one of the most compelling long-term themes adjacent to AI, with future adoption likely shaped by continued progress in hardware, energy availability, economics, and real-world reliability.

Closing Thoughts

The June 8 meeting reinforced several themes that surfaced repeatedly across otherwise very different discussions.

The first was that scarcity continues to create value. Whether the scarce asset is compute, power, semiconductors, land, data, distribution, public float, launch capacity, or skilled labor, many of the most attractive opportunities discussed throughout the evening were tied to assets that remain difficult to replicate. Across AI infrastructure, data centers, energy systems, healthcare, robotics, space, and financial markets, bottlenecks repeatedly emerged as some of the most important drivers of long-term value creation.

AI remained the connective thread linking much of the conversation. Topics ranged from Anthropic's work on recursive self-improvement and the implications of increasingly capable models to compute scarcity, power generation, cloud infrastructure, robotics, healthcare, autonomous systems, and orbital computing. A recurring observation was that AI is evolving beyond a software story and increasingly resembles a large-scale industrial buildout requiring enormous investments in physical infrastructure, energy, manufacturing, networking, and data.

SpaceX reflected many of these themes simultaneously. Discussion touched on launch systems, satellites, AI infrastructure, data centers, communications networks, robotics, and one of the most anticipated IPOs in modern market history. More broadly, the conversation highlighted how major technological shifts often create opportunities not only for the headline companies but also for the suppliers, infrastructure providers, component manufacturers, and ecosystem participants supporting them.

Another theme was the growing importance of market structure itself. Chinese micro-cap volatility, prediction markets, AI trading agents, crypto treasury companies, and tightly controlled IPO floats all illustrated how liquidity, positioning, incentives, and investor behavior can sometimes drive outcomes as much as underlying fundamentals. Understanding how markets function may be becoming just as important as understanding the businesses operating within them.

The discussion also reinforced the value of second-order thinking. The most obvious beneficiaries of a trend often receive the majority of investor attention, but many of the opportunities explored throughout the evening existed one or two layers deeper in the value chain. Infrastructure providers, data owners, semiconductor manufacturers, networking companies, power suppliers, automation platforms, and enabling technologies frequently emerged as some of the most interesting ways to gain exposure to broader secular trends.

Taken together, the evening's discussions pointed toward a common conclusion: many of the defining investment opportunities of the coming decade may be found not only in the

technologies attracting the most attention, but in the systems, infrastructure, and supporting ecosystems that make those technologies possible.

We look forward to reconvening at the next meeting of The Investing Group at the Union League Club of Chicago on June 22, 2026.