Key Takeaways
- Mike Winston has established a two-pronged AI infrastructure strategy through AI Infrastructure Acquisition Corp. (NYSE: AIIA) and Jet.AI (NASDAQ: JTAI), creating multiple pathways to participate in the expanding AI infrastructure market.
- Jet.AI owns a 49.5% economic interest in AIIA Sponsor Ltd., allowing JTAI shareholders to benefit from the sponsor economics if AIIA successfully completes a business combination.
- Alongside its SPAC investment, Jet.AI is developing data center infrastructure through its Convergence Compute joint venture, targeting approximately 1 gigawatt of planned capacity across multiple North American sites.
- The investment thesis is built around growing global demand for AI compute, increasing data center capital expenditure, and the persistent shortage of reliable power infrastructure needed to support next-generation AI workloads.
- While the strategy presents significant growth potential, investors should recognize the execution risks associated with SPAC timelines, infrastructure development, regulatory environments, and capital-intensive projects.
On Nov. 19, 2025, AI Infrastructure Acquisition Corp. (NYSE: AIIA) rang the opening bell at the New York Stock Exchange. The SPAC had raised $138 million in its initial public offering. The proceeds sat in trust, earmarked for a business combination in data centers or AI infrastructure. Mike Winston was its co-sponsor and CEO.
Eleven days earlier, Jet.AI (NASDAQ: JTAI), the company he founded and chairs, had disclosed that its stake in the AIIA sponsor entity was worth approximately $17 million on its balance sheet.
Two listed vehicles. One infrastructure thesis. And a structural link between them that neither press release spelled out.
Two Vehicles, One Infrastructure Thesis
The connection between JTAI and AIIA runs through the sponsor economics of the SPAC structure.
When AIIA completed its initial public offering in October 2025, it closed at 12 million units before exercising the overallotment option that brought the total to 13.8 million units and $138 million in gross proceeds. AIIA’s mandate is to identify and merge with a business in data center infrastructure or AI, which the company describes as “ship to grid.”
Jet.AI holds a 49.5% economic interest in AIIA Sponsor Ltd., the entity that controls the SPAC’s founder shares and warrants. This is ownership in the entity that receives the financial upside when a business combination closes, not a passive position in AIIA’s trust account. SPAC sponsors typically receive 20% of post-IPO equity as founder shares plus warrants exercisable at $11.50. Jet.AI’s 49.5% position means that if AIIA identifies and closes a qualifying deal, nearly half the sponsor economics flow back to JTAI shareholders.
As of Q1 2026, JTAI carried the AIIA stake at $17.23 million on its balance sheet. The company reported that AIIA was “actively engaged with several targets.”
The post-divestiture version of Jet.AI runs this AIIA position alongside its Convergence Compute joint venture with Consensus Core Technologies, a development program targeting data center campuses across a Midwestern location, a Maritime location, and a Manitoba site with 500 megawatts of confirmed natural gas supply. Three of four planned development milestones are complete. The combined capacity target across the three sites is 1 gigawatt.
JTAI is building data center infrastructure directly through Convergence Compute while AIIA assembles capital to acquire or merge with a business already operating in the same space. Mike Winston is the architect of both.
How Large the Cycle Is
Timing an infrastructure cycle correctly requires knowing how large the cycle actually is.
According to Dell’Oro Group research, global data center capital expenditure is approaching $1 trillion in 2026. The top four US hyperscalers (Amazon, Google, Meta, and Microsoft) are on track for combined capex of approximately $600 billion this year. Dell’Oro projects that curve extending to $1.7 trillion globally by 2030, with accelerated servers potentially accounting for two-thirds of data center spending by that point.
Power is the binding constraint. Data centers can source compute at scale; they cannot source enough electricity fast enough. Grid construction timelines for major interconnections in the United States commonly run five to seven years. That gap cannot close quickly. It is a structural supply problem with a long tail.
Jet.AI’s answer to this bottleneck involves aero-derivative engines: gas turbine technology adapted from aviation applications to power generation. These engines deploy faster than traditional grid connections and can be positioned adjacent to compute facilities. The Manitoba campus’s confirmed 500-megawatt natural gas supply is an early proof point.
“Given my background in real estate finance and telecom,” Winston said in an April 2026 interview, “it was a natural transition. Today, we’re extending that into power generation using aero-derivative engines, another area with strong underlying demand.”
A Telecom Career, an Infrastructure Cycle
Winston began his Wall Street career in 1999 at Credit Suisse First Boston, covering the telecommunications sector as an equity research analyst. His team’s research was ranked number one by Institutional Investor Magazine, a recognition that reflects analytical standing in the institutional investor community rather than output volume.
The timing of that career start matters. The late 1990s and early 2000s produced one of the largest infrastructure build cycles in modern history. Hundreds of billions of dollars flowed into fiber optic cable, switching equipment, and broadband networks based on projections of exponential traffic growth. Many of those projections were accurate: the demand was real. What created the crash was not that the infrastructure was unnecessary, but that capital committed faster than demand could absorb it, and the companies that survived were those with genuine cost and technology advantages, not simply those that moved first.
Winston observed that cycle from an analytical position that required him to model it in real time. He saw which companies survived and which failed, and the operational reasons underlying both outcomes. He saw how infrastructure spending becomes self-referential at scale: capacity attracts demand, demand attracts more capacity, and the companies that built genuine technical differentiation outperformed those that were merely early.
Data centers are not fiber networks. But the structural dynamic is familiar: a rapid capital commitment cycle, a power and grid bottleneck that capital cannot solve quickly, and a premium on locational and technological differentiation over speed-to-announce.
Winston’s view of compute is not framed as cyclical at all. “Societies that possess large amounts of compute and strong scientific and engineering cultures will ultimately dominate those that do not,” he said in an April 2026 interview. “The reason is straightforward: with sufficient compute, you can solve first-order scientific problems that have remained unanswered for decades or centuries. Once those are solved, you unlock entirely new classes of second-order problems.”
That is the argument of someone who does not think this build cycle ends. It is also the investment thesis that connects his telecom research training to his current position in AI infrastructure.
Background Behind the Bet
Winston founded Sutton View Capital in 2012, after departing Millennium Partners where he had co-managed a $1 billion merger arbitrage and event-driven book. At Sutton View, he has advised one of the largest academic endowments in the world and co-led successful activist litigation against the Dole Foods board of directors; that campaign secured a 35% increase in total consideration for shareholders.
The thread connecting merger arbitrage, corporate activism, and AI infrastructure is event-driven analysis: the discipline of identifying structural dislocations where the market’s current price diverges from the probable future value of an asset. In merger arbitrage, the dislocation is between the deal announcement price and the probability-weighted expected consideration. In activism, it is between current management’s capital allocation and what a competent board would do instead. In AI infrastructure, the argument is that the market has not yet fully priced the scale of the compute buildout, the duration of the power constraint, or the strategic weight that compute capacity will carry over the next several decades.
Sutton View’s founding, after two institutional platforms in Credit Suisse First Boston and Millennium Partners, was a deliberate choice. The CFA credential, the Institutional Investor ranking, the Columbia MBA: these are establishment markers. Building outside established platforms, on independent conviction, is a different posture. The combination defines the investor Winston projects: institutionally credentialed, independently positioned.
What the Position Requires to Work
AIIA has a finite window to close a business combination. Standard SPAC structure runs 18 to 24 months from IPO, with extension possible under the trust agreement. The company IPO’d in October 2025. No combination has been announced, though management has confirmed active engagement with several targets.
Convergence Compute has three milestones completed and one remaining. The Manitoba campus has confirmed natural gas supply. Construction, equipment procurement, and customer acquisition across three sites follow.
The risks are real. Infrastructure cycles can absorb more capital than they return. SPAC vehicles face structural pressures that operating companies do not: redemption windows, timeline constraints, and trust account mechanics that create their own deadline pressure. Power generation advantage is only durable as long as the regulatory and grid environment that creates the advantage remains stable.
What Winston has assembled is not a single-outcome bet. JTAI and AIIA are two entry points into the same thesis, structured so that a business combination in AIIA and an organic buildout through Convergence Compute can each succeed on its own. The 49.5% sponsor stake means JTAI shareholders participate in whichever path produces value first. The Convergence Compute milestones continue regardless of what happens in the SPAC market.
Two listed vehicles. One infrastructure thesis. Enough structural separation between them that not everything has to go right at once.
FAQs
Who is Mike Winston?
Mike Winston is an investor, entrepreneur, and infrastructure executive. He serves as CEO and co-sponsor of AI Infrastructure Acquisition Corp. (NYSE: AIIA) and is the founder and chairman of Jet.AI (NASDAQ: JTAI), where he focuses on AI infrastructure and data center development.
What is AI Infrastructure Acquisition Corp. (AIIA)?
AIIA is a special purpose acquisition company (SPAC) formed to identify and merge with businesses operating in AI infrastructure, data centers, and related power-generation sectors.
How is Jet.AI connected to AIIA?
Jet.AI owns a 49.5% economic interest in AIIA Sponsor Ltd., the entity that holds the SPAC’s founder shares and warrants. This structure allows Jet.AI shareholders to participate in the financial upside if AIIA successfully completes a qualifying business combination.
Why is power infrastructure so important for AI data centers?
Modern AI workloads require enormous amounts of electricity. As demand for AI computing continues to rise, the availability of power has become one of the primary constraints on new data center development, making alternative power generation and energy infrastructure increasingly valuable.
What are the primary risks associated with this investment strategy?
Key risks include SPAC execution deadlines, business combination uncertainty, infrastructure construction delays, regulatory changes, power supply challenges, and the capital-intensive nature of large-scale AI infrastructure projects.