Rational Choices Amid the AI Investment Boom: North American Tech Capital Redefining Growth Logic

When Growth Becomes the 'New Normal': Speed and Concerns in the AI Market

In 2025, the capital frenzy in the AI field shows no signs of braking. At TechCrunch's StrictlyVC event, a conversation between M13 co-founder Carter Reum and Basis Set Ventures partner Chang Xu revealed the current market's "paradoxical prosperity": on one hand, the valuation growth rate of AI startups is unprecedented; on the other hand, the actual revenue of these companies is also expanding rapidly. For example, ChatGPT increased its revenue from $1 million to $40 billion within six months; some startup teams of just 20 people saw annual revenue jump from $10 million to $70 million.

This growth rate puts traditional business models to shame, but it also raises deep concerns among investors—is this truly the explosion of a technological revolution, or another bout of irrational exuberance?

Why It Happens: Resonance of Technology, Capital, and Platform Effects

The "lightning-fast growth" of the AI market is no accident. Over the past two years, breakthroughs in large language models and generative AI have moved AI from the lab to large-scale commercialization. Unlike the era of cloud computing or smartphones, the accelerator of this round of technological iteration is the "platform effect": improvements in underlying model capabilities can instantly empower thousands of application scenarios. Each time leading platforms like OpenAI, Google, and Anthropic release a new feature, they may reshape the entire industry chain.

This effect also significantly lowers the barriers to entrepreneurship. Developers no longer need to build their own models; they can develop products based on mature APIs. Market demand is far more voracious than supply—companies' eagerness for AI to reduce costs and increase efficiency means that as long as a product offers basic value, it can quickly acquire paying users.

Who Benefits, Who Bears Pressure: The 'Invisible Wall' of Tech Giants

However, the "universality" of opportunity does not mean all participants will survive. Carter Reum pointed out that this AI cycle is different from the past: innovators are no longer just competing with their peers, but directly against the world's ten most valuable tech giants. Apple, Google, and Microsoft not only have massive capital and talent reserves but also possess vast user data and complete ecosystems.

For small startups, being "eaten up" or "squeezed out" is the norm. Reum emphasized that "moats" are more critical than ever—startups must focus on areas where tech giants have yet to penetrate, especially industries with complex regulations or high data barriers, such as healthcare, government services, and law. In these fields, compliance barriers and industry knowledge can form effective defenses.

Conversely, AI applications that lack differentiation and simply rely on general-purpose model capabilities will face immense pressure. Once platform providers roll out similar features, these startups may instantly lose their customers. For example, if OpenAI embeds document editing or data analysis features into ChatGPT, third-party tools relying on those functions would lose their reason to exist.## Significance for Investors: From "Betting on Trends" to "Betting on Structures"

In the face of a rapidly changing market, investor strategies are also evolving. Basis Set Ventures adopts a "dual-track" investment framework:

• AI Infrastructure (Under): Focus on databases, version control, deployment tools, etc. These tools were originally designed for humans and now need to be restructured to adapt to AI agents. The infrastructure layer has high technical barriers and platform stickiness, forming a more durable moat.
• AI Applications (On Top): Select products with long-term differentiation capabilities, especially those with unique protection in technology or data.

This strategy reflects investors' wariness of a "bubble." Currently, many AI companies appear overvalued, but their actual revenue supports the prices. However, calculations based purely on financial models may fail—because the leading foundation model companies could release new features at any time, rendering dozens of startups' products obsolete overnight.

Therefore, the essence of investing has shifted from "finding the next unicorn" to "identifying structural safe zones." Capital is flowing into areas that can withstand platform shocks: complex industry data, highly regulated processes, and enterprise customer relationships that require years of trust to build.

Industry Chain Restructuring: New Flows of North American Tech Capital

The AI boom is reshaping the tech capital landscape in North America. Traditional Silicon Valley venture capital is increasingly investing in the infrastructure layer, while funds previously concentrated on the application layer are diverging. At the same time, government involvement in AI regulation (such as the EU AI Act and US executive orders) is making compliance capability a new valuation dimension.

Notably, this trend also affects ecosystems outside North America. The original text mentions emerging markets like Uzbekistan, which can avoid direct competition with giants and leverage their unique industry data or regulated sectors (e.g., healthcare, government) to build AI solutions. For North America, this means AI investment geography may spread from Silicon Valley to other regions with specialized data or industry characteristics—such as medical AI in Boston, government AI in Washington, and energy AI in Texas.

Long-term Trends: Three Key Judgments for the Next 3–5 Years

1. Infrastructure Oligopolization: The AI model layer will form 2–3 dominant players. Whoever controls the more powerful foundation model will define the standards for downstream applications. The infrastructure layer (especially computing power and toolchains) will also see consolidation. 2. Application Layer "Slimming" and Differentiation: General-purpose AI applications will be absorbed by platforms, while AI companies deeply rooted in vertical industries (e.g., legal, healthcare, defense) will command a premium through data barriers and compliance capabilities. 3. Capital Efficiency First: Investors will no longer chase "burn money for growth" models, instead demanding clear commercialization paths and technical moats. AI companies that can prove sustainable cash flow will gain more favor, while startups relying solely on funding will face a winter.

ConclusionThe "unprecedented" nature of the AI market is both an opportunity and a trap. At a time when North American tech capital is being redefined, investment logic must shift from a worship of speed to a pursuit of structural stability. Only those players who can simultaneously understand technological evolution, platform competition, and the regulatory environment can truly prevail in this transformation.