The AI market isn't a monolithic bubble but rather three distinct layers, each with its own survival strategy and timeline. These segments exhibit different vulnerabilities and resilience, shaping the future landscape of artificial intelligence.
At the top is the wrapper layer, where companies resell AI capabilities without building core models themselves. They often charge premium prices—$49 or $50 for access that feels like ChatGPT—but their business models are fragile. Some have grown rapidly, with one reportedly reaching $42 million in annual recurring revenue in its first year. However, these companies lack defensibility. Without owning data, workflows, or deep integrations, customers can switch to competitors—or directly to ChatGPT—in minutes without cost. The exception is Cursor, which has built strong network effects through developer workflows, but most wrapper businesses won't survive the coming consolidation.
The foundation model layer—companies like OpenAI, Anthropic, Mistral—occupies a more stable but still precarious position. These firms possess genuine technological moats: model training expertise, compute access, and performance advantages. However, as baseline capabilities converge, the competitive edge will shift to inference optimization and systems engineering. Winners in this space will be those who can scale memory walls through innovations like extended KV cache architectures, achieve superior token throughput, and deliver faster time-to-first-token. Technical breakthroughs in memory management and infrastructure efficiency will determine which frontier labs survive consolidation.
Another concern is the circular nature of investments, such as Nvidia pumping $100 billion into OpenAI to bankroll data centers, which are then filled with Nvidia's chips. This could artificially inflate AI demand, creating a feedback loop that distorts market dynamics. Despite these risks, foundation model companies have massive capital backing and strategic partnerships with major cloud providers and enterprises. Consolidation is expected between 2026 and 2028, with 2 to 3 dominant players emerging while smaller model providers are acquired or shuttered.
The infrastructure layer—Nvidia, data centers, cloud providers, memory systems, and AI-optimized storage—is the least bubbly part of the AI boom. While global AI capital expenditures and venture capital investments exceed $600 billion in 2025, infrastructure retains value regardless of which specific applications succeed.
Nvidia's Q3 fiscal year 2025 revenue hit about $57 billion, up 22% quarter-over-quarter and 62% year-over-year, with the data center division alone generating roughly $51.2 billion. This represents real demand from companies making genuine infrastructure investments.
The current AI boom won't end with one dramatic crash. Instead, we'll see a cascade of failures beginning with the most vulnerable companies. Hundreds of AI startups with thin differentiation will shut down or sell for pennies on the dollar. Foundation model consolidation will follow as performance converges and only the best-capitalized players survive.
Infrastructure spending will normalize but remain elevated, with some data centers sitting partially empty for a few years before eventually filling as AI workloads genuinely expand.
The most significant risk isn't being a wrapper—it's staying one. If you own the experience the user operates in, you own the user. Building in the application layer requires moving upstack immediately, owning the workflow before and after the AI interaction.
Winning AI businesses aren't just software companies—they're distribution companies. Understanding which layer you're operating in and which bubble you might be caught in is the difference between becoming the next casualty and building something that survives the shakeout.
The AI revolution is real, but not every company riding the wave will make it to shore.
