Investing in Private Growth

Vertical AI companies are experiencing especially rapid growth in areas such as legal, finance, sales preparation, and selected medical fields. In these categories, AI does not merely reduce costs — it strengthens the business model itself by directly driving revenue and expanding the addressable workflow.

The "Cursor-for" concept is a useful illustration. These products are not simple interfaces to a single model. Instead, they function as a full execution layer that: handles context engineering, orchestrates multiple LLM calls through a Directed Acyclic Graph (DAG), balances cost and output quality dynamically, provides a specialized workflow-native UI, and introduces an "autonomy slider" that controls how much work is delegated to the system.

Looking into 2026, we expect the next step to be multiplayer mode. Most vertical work is inherently multi-party — lawyers, analysts, doctors, operators rarely work alone. As these tools evolve to support collaboration, shared context, and handoffs, we may finally see the network effects that have largely eluded AI applications so far.

Established technology companies are actively trying to capture value from what increasingly looks like a new industrial revolution driven by AI and robotics. In public markets, investors will closely watch how Fortune 500 companies attempt to translate AI adoption into real productivity and margin improvement.

Why now? The idea that AI could replace expensive human labor has been discussed for years. The difference in 2026 is execution. Enterprises are beginning to move away from isolated AI tools toward multi-agent systems that behave more like coordinated teams, while physical AI enables deployment in manufacturing, logistics, and operations.

One of the clearest competitive arenas is e-commerce. Google, Shopify, OpenAI, Stripe, and others are racing to define "agentic commerce." At the same time, major uncertainties remain. It is unclear whether consumers will meaningfully adopt chatbot-driven purchasing when existing shopping habits are deeply ingrained.

For AI systems to operate predictably, scalably, and safely during this "year of execution," demand will grow for what can be described as the "shovels and shields" of AI. This includes next-generation infrastructure, orchestration layers, cybersecurity, and identity solutions. Each should be viewed not as incremental upgrades, but as a new generation of its category.

A clear example is the resurgence of neoclouds. The first wave of neocloud providers — including CoreWeave, Crusoe, and Lambda — gained traction by offering faster, cheaper, and more AI-optimized compute than hyperscalers, largely by securing early access to Nvidia Hopper chips and avoiding internal competition from in-house AI teams.

The current resurgence is driven by a different dynamic. As the original neoclouds hit capacity limits and reprioritize customers, they are quietly declining new deals. This creates an opening for a new cohort of players to step in and capture unmet demand.

The coming year is also likely to bring significant developments across emerging technologies — including robotics, autonomous driving, nuclear energy, quantum computing, and defense. These areas are expected to dominate headlines, driven by large government contracts and increasingly favorable regulatory frameworks, making them prime targets for new waves of hype-driven investment.

The hottest segment within this group is physical AI. Earlier robotics waves were defined by narrow, pre-programmed machines designed for limited tasks in controlled environments. Today's AI advances enable general-purpose robots across multiple form factors, capable of learning, adapting, and operating in complex, real-world settings.

This shift is forcing a full rewrite of the robotics stack. The competitive battlefield is global, with rival ecosystems forming across regions and supply chains.

Several forces are converging in consumer AI: falling inference costs, expanding multimodal context windows, cheap code generation, and the potential emergence of new software distribution ecosystems. Together, these dynamics are creating fertile ground for a new wave of consumer AI products that are personalized, multimodal, and no longer limited to a chat window.

Much of the creative experimentation will happen at the UI layer. We are already seeing the rise of AI-native creativity tools — covering generative voices, music, images, and video. These building blocks now form a complete toolkit for AI-driven storytelling, making content creation one of the clearest consumer use cases for AI.

Pre-IPO investing — a popular strategy in 2020 and 2021 — is showing signs of a comeback. Investing in companies shortly before their expected public offering can offer relatively fast liquidity, allowing capital to be recycled more quickly into subsequent opportunities.

At the same time, strategic mega-deals ($5B+) are returning, marking a shift away from the take-private–dominated environment of recent years and improving exit visibility across the private market.