AI's Trillion-Dollar Opportunity Is Bigger Than Software

Based on Sequoia's AI Ascent 2025 keynote.

For years, the biggest technology shifts were understood through software.

First came desktop software. Then cloud. Then mobile. Each wave created massive companies by changing how software was built, distributed, and monetized.

AI may be different.

The argument from Sequoia's AI Ascent 2025 keynote is that AI is not just attacking software budgets. It is also attacking labor budgets. That matters because the real opportunity may be much larger than previous platform shifts. If cloud expanded software, AI may expand what software can replace, coordinate, and execute.

That is a much bigger market.

Why AI may move faster than previous technology shifts

One reason this shift feels different is speed.

Previous technology transitions often needed time to build awareness, distribution, and connectivity. AI started with all three already in place. The world was immediately paying attention. Billions of people were already online. Distribution channels already existed. The result is that AI adoption has moved much faster than earlier platform changes.

This matters because it compresses time.

Companies may not have years to adapt to a new way of being discovered, compared, and chosen. The infrastructure that wins in an AI-native market may be established much earlier than many incumbents expect.

The real value may still be built in the application layer

Another important point from the keynote is where value is likely to accrue.

Even with increasingly capable foundation models, Sequoia's view is that a great deal of value will still be created in the application layer. That is especially true where the problem is vertical, operational, messy, and tied to real customer workflows rather than generic model access.

In other words, the winners are unlikely to be the companies that just wrap a model.

They are more likely to be the companies that take a real customer problem, structure the workflow, handle the constraints, and turn capability into a usable product or service.

That is an important distinction. Model capability alone does not solve real-world execution. Application design does.

From tools to copilots to autopilots

One of the most important ideas in the keynote is that AI products are evolving.

They begin as tools. Then they become copilots. Eventually, in some categories, they become autopilots.

A tool helps a person do the work. A copilot assists a professional doing the work. An autopilot takes responsibility for larger parts of the workflow.

That shift matters because it changes where budgets come from.

When software moves closer to execution, it stops competing only for software spend. It starts competing for spend that previously went to labor, coordination, outsourced services, and manual operations.

That is where the market can become much larger.

The next wave is not just AI apps. It is an agent economy.

The keynote also points to a bigger structural shift: the emergence of an agent economy.

In that world, AI agents do not just answer questions. They communicate, make decisions, transfer resources, complete transactions, and develop persistent working relationships with people and systems.

That is a meaningful step beyond chat interfaces.

It suggests a future where agents are not only part of the user interface, but part of the execution layer of the economy itself. For that to work, markets will need better identity, better protocols, better trust systems, and better infrastructure for machine-mediated transactions.

This is where AI becomes less about content generation and more about economic coordination.

Why this matters for services

This shift becomes even more important in services.

Product commerce is already relatively structured. Products can often be indexed, priced, stocked, compared, and purchased through static flows. Services are different.

Services are variable. They are often negotiated. Availability matters. Timing matters. Capacity matters. Scope is often unclear at the start. And closing the job usually requires coordination.

That means services are one of the categories where simple discovery is not enough.

If AI is going to matter in services, it needs to do more than help users search. It needs to help users and agents move from intent to real execution.

What this means for Aune

This is exactly where Aune fits.

Aune is built around the idea that real-world services break product-centric systems. Search, listings, and marketplaces can show options, but they do not solve the harder problem: which provider can actually take the job, under what constraints, at what price, in what time window, with what fallback if something changes.

That is why Aune is not just another discovery layer.

Aune is building transaction infrastructure for agent-driven service commerce. The role is to help users and AI agents move from a request to a feasible, negotiated, committed booking. That means handling discovery, matching, negotiation, commitment management, and learning from real outcomes over time.

In a world where AI is moving from tools to autopilots, and from chat to agent economy, that matters.

Because the winner in services may not be the one that helps users browse providers more efficiently. It may be the one that helps agents and users actually get the job done.

Why this creates a real opportunity

If Sequoia is right, then the biggest AI opportunities will not only be found in foundation models or horizontal assistants. They will also emerge in application-layer companies that own difficult workflows in large categories.

Services fit that pattern well.

They are fragmented. They are operationally messy. They involve real constraints. And they have historically required manual coordination.

That creates room for a new layer of infrastructure.

For Aune, that means the opportunity is larger than building a better directory of service providers. The opportunity is to become part of the system that lets AI agents and users transact in real-world services with confidence.

That is a much bigger ambition.

Final thought

The AI wave may be bigger than cloud not just because models are powerful, but because AI reaches beyond software and into execution.

As that happens, value may shift toward the companies that can turn messy real-world demand into structured outcomes.

In products, that might mean smarter commerce.

In services, it may require an entirely new transaction layer.

That is the shift Aune is building for.