AI’s Second Act: From Suppliers to “Appliers”

Why value is shifting from AI infrastructure to the businesses that integrate and monetize it.

ChatGPT has only been on the scene since 2022. Since then, the world has raced to build the “picks and shovels”—the chips, power, and data‑center infrastructure that make AI possible. Now the story is shifting: the appliers who weave AI into products and workflows appear positioned to define the next decade.

Since launch, the AI narrative has largely revolved around the companies enabling the build‑out of infrastructure—semiconductors, electrical equipment, power generation, and data‑center construction. We believe the opportunity in these suppliers will continue. 

But history shows that the baton eventually passes from the suppliers to the appliers—businesses that embed new technology into products, processes, and, in some cases, entirely new categories. The late 1990s internet cycle offers a useful framing: early value accumulated to the companies laying the foundation (fiber optics, networking gear, servers), where demand was driven by capex cycles rather than true end‑market adoption. Over time, however, the biggest winners were those that integrated the internet into their models to scale, serve customers better, and pioneer new business models. Many of today’s mega‑caps—including Amazon and Google—emerged from that transition.

We believe a similar dynamic is unfolding with AI. Infrastructure remains essential, but value creation is increasingly shifting from the “picks and shovels” providers to the appliers—businesses that integrate and monetize AI within their products. To illustrate this shift, Figure 1 provides a conceptual view of how we see value migrating from suppliers to appliers across the AI ecosystem. 

Figure 1: Shifting Value Capture in the AI Ecosystem: From Suppliers -> Appliers

Image

Health Care

We think Health Care is among the sectors most ready for AI-driven change. Health Care faces high administrative costs, clinician burnout and shortages, fragmented data, and inconsistent patient experiences. AI can lower administrative burdens for busy clinicians and improve triage decisions by helping identify which patients need higher-acuity care sooner. 

Beyond services, AI is also reshaping biotech and pharma. Drug discovery and R&D have historically been slow and expensive. AI can accelerate early-stage screening by rapidly evaluating potential compounds and modeling interactions before lab testing. In clinical trials, AI can help identify eligible patients faster and support real-time monitoring, potentially shortening timelines and improving productivity.

Consumer

For consumers, the key is to distinguish surface-level AI features from genuine integration into the customer experience.  Companies with strong first-party data can leverage AI to personalize discovery and recommendations, boost loyalty, and reduce friction. Even returns can be improved if AI clarifies product details, refines sizing, and predicts return risk. This not only lowers high operational costs but also enhances the overall customer experience.

AI can also help brands respond faster to shifting preferences by enhancing demand forecasting, inventory management, and supply chain decisions. Over time, AI might open up entirely new consumer categories. Humanoid robotics is one emerging example at the intersection of consumer and industrial applications. Although estimates of the market potential vary widely, a quick look at Tesla’s current valuation provides some insight into the market’s excitement for humanoid robotics over the next decade. 

Industrials

Industrials often operate with high fixed costs and long-lived assets, which makes improved uptime and efficiency especially valuable. Highly variable capex and economic cycles govern the profitability of many of these companies, and this lack of predictability is why many investors ascribe a below-market multiple. One of the most direct applications is predictive maintenance, shifting from scheduled maintenance to condition-based insights that anticipate failures before they occur. For airlines, railroads, utilities, and factories, this can reduce unplanned outages, extend asset life, and improve working capital efficiency. AI also supports autonomy and robotics where labor is scarce, safety matters, and consistency is critical, such as logistics, warehousing, agriculture, mining, and ports. A third area is manufacturing yield and quality optimization. AI can detect defects earlier, optimize process parameters in real time, reducing scrap and rework, and help avoid costly quality failures. 

Closing Thoughts

The next decade of AI leadership likely won’t be defined by the companies building the tools. It is expected to be defined by the companies using them. Those that weave AI into mission‑critical workflows, unlock new customer value, and enhance unit economics stand to reshape industries. If past technological shifts are any guide, the enduring winners may be the appliers who translate capability into competitive advantage. For investors, spotting where AI creates structural differentiation, not just short‑term hype, will be key to capturing the next wave of long‑term value creation.

Download the paper