Why the "service as software" with Generative AI is a trillion-dollar market.

Sequoia, the prominent venture capital firm in Silicon Valley, analyzed in its new article 'Generative AI’s Act o1' how it is "quite feasible" that startups could "outperform global systems and corporate IT integrators" by focusing on GenAI at the application layer.

1. The transition to the cloud has brought software as a service (SaaS). The transition to AI consists of creating applications that automate and amplify knowledge work, "service as software," a market much larger that is worth trillions (see chart).

2. Competition at the base layer (infrastructure and models) is stabilizing, led by Microsoft/OpenAI, AWS/Anthropic, Meta, and Google/DeepMind, due to high capital requirements.

3. GenAI is evolving from "thinking fast" (pre-trained responses) to "thinking slow" (OpenAI's reasoning in inference), unlocking the potential for "revolutionary applications" of AI/agentic in consumer and enterprise markets.

4. Two years ago, many said that companies in the application layer were "just a wrapper for GPT". Today, those "wrappers" have become one of the best ways to create lasting value.

5. What started as wrappers has evolved into "cognitive architectures": multiple models, routing mechanisms, vector/graph databases for RAG, compliance guardrails, and application logic that reflects human reasoning in workflows.

6. Emerging examples: Harvey's AI lawyer, Glean's AI work assistant, Factory's AI software engineer, Abridge's AI medical scribe, XBOW's AI pentester, Sierra's AI customer support agent.

7. However, transforming GenAI capabilities into reliable comprehensive enterprise applications still requires significant engineering effort and business expertise.

8. Established companies must evolve or risk being overtaken by new firms that master the complexities of knowledge work and manage to reach customers.

9. Similar to the emergence of 20 companies worth over 1 billion dollars in mobile and cloud applications, Sequoia expects a comparable opportunity with AI.

Why it matters:

There is great potential for reinvention and business disruption. Innovation at the application layer will play a key role in defining the next generation of industry leaders.

The choice is clear: disrupt or be disrupted.