Bret Taylor of Sierra on AI agents, outcome-based pricing, and the OpenAI board

Agentic AI is fundamentally reshaping software engineering and business operations by shifting the focus from manual coding to the creation of robust "harnesses" that guide autonomous agents. This evolution favors structured, process-oriented environments where AI can reason through complex tasks, such as customer service or supply chain onboarding, rather than relying on generic multi-agent architectures. By treating business processes as code-like structures, companies can achieve higher fidelity outcomes and significant efficiency gains. Sierra’s implementation of AI agents across healthcare and finance illustrates this shift, as agents now manage end-to-end customer journeys—including voice interactions—at a fraction of the cost of traditional human-led support. Ultimately, the most valuable future software will likely be defined by these optimized, agent-compatible processes rather than just the underlying databases, marking a transition toward outcome-based business models that prioritize efficiency and direct value creation.