Sequoia Capital 2026 AI Trend Outlook: Computing Power Development Slows, but AI Startup Growth Momentum Remains

Sequoia Capital (Sequoia Capital) Partner David Cahn pointed out in the 2026 AI Trends Outlook report that the market is currently in a transition phase where investment sentiment is cooling, but the fundamentals of AI continue to advance. Moving into 2026, AI development will proceed along two parallel paths: one involves delays in data center construction and AGI timelines, and the other involves continued expansion of AI applications, especially the growth momentum of AI startups, which has not slowed despite market cooling.

Investment scale continues to expand, but AI revenue reality still lags

Cahn stated that before 2026, large tech companies’ capital expenditure demands for AI remain strong. Google and Meta continue to increase investments, while Microsoft and Amazon have made slight adjustments, but their overall long-term strategic directions remain unchanged. However, the supply chain has begun to show signs of fatigue, with some suppliers adopting a cautious attitude toward the sustainability of end-demand.

From a revenue perspective, the actual revenue-generating scale of the AI industry still only reaches hundreds of billions of dollars, forming a stark contrast with the future five-year investments in data centers and energy infrastructure, which are expected to reach trillions of dollars. In terms of applications, products that have achieved scale are still mainly in programming tools and models like ChatGPT. Most large enterprises face integration difficulties and limited results when deploying AI internally.

Rapid demand growth, supply chain becomes the first stress test

Cahn believes that the first structural pressure in 2026 will come from the continued rise in compute demand, but the reality is that supply chain expansion cannot keep pace immediately.

In advanced manufacturing processes, TSMC and ASML hold highly concentrated, near-monopoly industry positions. Their capacity expansion pace is relatively cautious and cannot be easily accelerated externally. Against the backdrop of rising demand for AI chips, this expansion pace may turn into actual capacity constraints by 2026.

Delays in construction and labor shortages gradually pose risks

In addition to wafer manufacturing, Cahn also pointed out the risks hidden in the latter stages of data center construction. As construction enters its final phase, key industrial equipment such as generators and cooling systems, as well as technically experienced personnel, may become bottlenecks affecting progress.

Since most AI companies share the same supply system, any delay in a single link could force the entire construction schedule to be postponed. Sequoia notes that 2026 will be a critical year for the market to comprehensively assess whether these industrial supply and labor bottlenecks are truly ready.

Construction cycle enters a showdown phase, AGI timeline shifts later

Cahn indicated that an AI data center typically takes about two years to complete. With a large number of projects starting in 2024 and construction investments beginning to reflect in economic data in 2025, 2026 will officially enter the results verification phase.

If major cloud providers start stockpiling AI chips without immediate deployment, it will be seen by the market as a significant signal of delay. Meanwhile, another delay stems from revisions to AGI development expectations. The previously popular “2027 AGI” forecast has been gradually pushed back since mid-2025. The new mainstream view leans toward the earliest realization of AGI being in the 2030s.

AI adoption continues to accelerate, startups become the main beneficiaries

Compared to the uncertainties surrounding hardware construction and AGI timelines, Cahn sees the trend of AI adoption itself as relatively clear. Even as market enthusiasm cools, the deployment of AI in practical applications continues to accelerate. Excellent startups are rapidly growing from zero to hundreds of millions of dollars in revenue, and from 2026 onwards, new startups with billion-dollar revenues are expected to emerge.

Cahn also observed that top AI startups demonstrate high operational efficiency, with some companies generating over one million dollars in annual revenue per employee, heavily utilizing AI agents to optimize internal processes, creating a self-reinforcing operational flywheel. In contrast, large enterprises face organizational and integration resistance when deploying AI internally, resulting in limited effectiveness. This creates more development space for startups focused on specific scenarios and mature products.

This article, Sequoia Capital 2026 AI Trends Outlook: Computing Power Construction Slows, but AI Startup Growth Momentum Remains Strong, first appeared on Chain News ABMedia.