Where AI’s Infrastructure Boom Is Creating New Opportunities

Preface

Artificial intelligence is often discussed in terms of models, chips, and applications. Yet behind every expanding AI service is a physical infrastructure that must supply power, connect equipment, house computing systems, and keep them operating. As demand grows, these foundational needs are becoming a defining part of the AI economy. At TechCrunch Disrupt 2026, Ben Longmier, CEO of Ambrosia Energy, and Bill Thayer, SVP and Head of Datacenter Solutions at Bloom Energy, will discuss where this infrastructure boom is creating opportunities. Their session, “Where the AI Infrastructure Boom Is Creating Winners,” will take place on the Smart Systems Stage. This article explores the core questions behind that conversation: which constraints matter, how lasting markets may emerge, and what founders, investors, and technology leaders can learn from the shift of AI into the physical world.

Lazy bag

AI growth depends on more than software. It also requires power generation, grid connections, data centers, cooling, electrical equipment, and reliable operating systems. The gap between rising demand for computing and the pace of infrastructure delivery may create opportunities for new businesses. The central challenge is distinguishing temporary shortages from persistent needs that can support durable markets. At TechCrunch Disrupt 2026, Ben Longmier and Bill Thayer will examine how these pressures are changing the market and where founders should look. The next significant AI opportunity may be in the systems that enable AI, rather than in an AI application itself.

Main Body

Artificial intelligence may be developed through software, but scaling it depends on a much broader set of resources. Models must run on computing equipment, applications need dependable access to that equipment, and the facilities hosting it require substantial power and supporting systems. As AI adoption expands, the question is no longer only how quickly software can improve. It is also whether the physical infrastructure can be built, connected, and operated at the pace that demand requires.

This distinction helps explain why the AI infrastructure boom is attracting attention beyond the companies that build models or chips. A functioning AI ecosystem relies on an interconnected stack that includes power generation, grid connections, data centers, cooling, electrical equipment, and systems for monitoring and maintaining operations. Each layer contributes to the availability, capacity, and reliability of computing. A limitation in one part of the stack can affect the usefulness of investments elsewhere.

At TechCrunch Disrupt 2026, Ben Longmier, CEO of Ambrosia Energy, and Bill Thayer, SVP and Head of Datacenter Solutions at Bloom Energy, will discuss this changing landscape at the Smart Systems Stage. Their session, titled “Where the AI Infrastructure Boom Is Creating Winners,” is intended to consider where enduring business opportunities may arise as AI places new demands on energy and data infrastructure. Bringing perspectives from energy and data center solutions can help make the discussion relevant to the practical dependencies behind AI growth.

One central issue is the pace mismatch between computing demand and infrastructure delivery. When the need for computing capacity increases faster than power or facilities can be made available, businesses encounter constraints that can delay projects or limit expansion. Those constraints are not automatically opportunities, however. A shortage may be temporary, tied to a particular location or moment, or it may reflect a lasting need that calls for new products, services, or business models. Determining which situation applies is essential for anyone deciding where to invest time or capital.

The discussion will therefore consider how to separate short-term pressure from structural change. If a bottleneck persists as AI usage grows, it may create space for a new category of companies. If it disappears as projects come online or conditions change, businesses built around that constraint may find demand less durable than expected. Founders and investors need to understand the source of a limitation, the customers affected by it, and whether a solution can be delivered at a scale and cost that make it commercially sustainable.

Energy is one of the most visible elements in this physical stack. Computing facilities need a dependable power supply, and an expansion in AI workloads can make power availability a planning concern. Yet energy is only one part of the challenge. Connecting a facility to the grid, equipping it to distribute electricity, and coordinating the systems that support its operation all matter as well. The opportunity may therefore appear in a number of places across the energy and electrical ecosystem, rather than in a single, obvious product category.

Data centers are another critical part of the picture. They provide the physical environment in which computing equipment operates, and their design must account for power, cooling, space, and the systems needed to maintain reliable service. As requirements change, businesses may need new ways to plan, build, manage, and support these facilities. Infrastructure software may also play a role by helping operators coordinate complex systems, though the specific opportunities depend on the operational problems customers are trying to solve.

Cooling illustrates how one infrastructure need can connect to several others. Equipment that performs intensive computing generates heat, so facilities must manage temperature as part of maintaining operations. Cooling requirements interact with facility design, power use, equipment choices, and operating procedures. This means solutions are not necessarily isolated: an improvement in one area may affect costs or performance elsewhere. Entrepreneurs considering this space need to understand how a proposed product fits into the larger system and whether it addresses a problem that customers prioritize.

Electrical equipment and grid technology may likewise become more important as demand for computing infrastructure grows. A facility’s ability to use power depends on more than the supply of energy; it also depends on the connections and equipment that deliver it where it is needed. The resulting challenges can create demand for solutions that help projects move forward or systems operate effectively. However, identifying a promising category requires more than observing rising demand. Founders must establish who the buyer is, what constraint the product resolves, and whether that need is likely to persist.

For founders, the main lesson is to look beyond the familiar software layer. The next business opportunity connected to AI might not be an AI application. It could emerge in energy, data centers, cooling, grid technology, electrical equipment, or infrastructure software. It might also take shape in a category that has not yet acquired a widely recognized name. The key is to begin with a specific physical-world need and determine whether it is becoming more important as AI scales.

Investors face a related challenge. Rising interest in AI infrastructure does not mean that every company serving the sector will build a durable business. Investors need to assess whether demand is tied to a temporary shortage or to a lasting change in how AI systems are built and operated. They also need to consider where value can be captured across a complex stack and which solutions can become difficult to replace. Understanding the underlying infrastructure may help distinguish companies with a clear role from those relying primarily on broad expectations about growth.

Technology leaders can benefit from the same perspective when planning for the next phase of AI adoption. Decisions about computing capacity increasingly depend on physical systems, including available power, facility readiness, cooling, and dependable operations. A clearer view of those dependencies can help organizations understand what must be in place before AI services can expand. It can also make infrastructure a strategic consideration rather than an afterthought.

Infrastructure booms do not distribute benefits evenly. Some parts of the stack may attract sustained investment because they address persistent needs; others may see short-lived demand or face barriers that limit growth. A useful analysis asks where constraints are appearing, how the market is responding, and whether a proposed solution can support a durable business. The session at TechCrunch Disrupt 2026 will bring these questions into a conversation spanning energy, data center solutions, and market intelligence.

That broader view matters because AI’s physical requirements are connected. Power availability can affect data center development, while the facility’s equipment and operating systems influence how computing capacity is deployed. Cooling and electrical systems add further considerations. Evaluating these factors together can reveal opportunities that are easy to miss when AI is viewed only through models, chips, and applications.

The central idea is that AI creates new demands on the physical world, and those demands need solutions. Some of the next markets may be found in established sectors adapting to new requirements; others may emerge where existing categories do not adequately describe the problem. For entrepreneurs, investors, and technology leaders, the task is to identify real constraints, understand who experiences them, and evaluate whether an answer can meet a lasting need.

Ben Longmier and Bill Thayer will explore these themes at the Smart Systems Stage at TechCrunch Disrupt 2026. The event features 250+ speakers leading 200+ sessions. Moscone West opens on October 13 at 8 a.m. PT. The event information states that attendees can save up to $100 before doors open and receive 50% off a second pass of the same ticket type when bringing a co-founder, partner, colleague, or peer.

Key Insights Table