Aaron Levie explores the future of enterprise software in a landscape increasingly influenced by AI agents. He emphasizes that companies invest in software to streamline their core processes, which provide essential structure for operations. For example, a large car manufacturer relies on an ERP system to manage vast financial transactions with high reliability. Levie introduces the concept of “core” versus “context” in enterprise software. Core elements define a company’s unique identity, while context includes necessary systems like payroll and IT support that don’t differentiate firms but are critical for smooth operation. As AI agents proliferate in enterprises, Levie argues that they will not replace traditional software but rather enhance it. He compares software to machinery in a factory and AI to the operators. Deterministic software ensures consistent performance, while non-deterministic systems allow AI to handle tasks that require flexibility, like generating sales presentations or analyzing documents. This synergy means that the value of traditional software will actually increase, as it will provide the framework within which AI operates. Levie highlights a significant shift in budget dynamics for software companies. Traditionally, software budgets are capped at 3-7% of a company’s revenue, limiting the growth of various tech markets. With AI agents taking on more work, the budget considerations will expand beyond just technology costs to encompass the overall spending on tasks these agents can perform. For instance, AI in legal services could tap into a $400 billion market, vastly outpacing previous software limitations. This trend is evident in coding and other knowledge work sectors, where AI-generated revenue far exceeds that of traditional tools. The challenge for existing software companies lies in adapting to this rapid evolution. Levie references Clayton Christensen's Innovator's Dilemma, suggesting that incumbents may struggle to pivot if new technologies don’t align with their existing business models. While some will adapt and thrive, others may fall behind as the market shifts. The landscape is shifting quickly, and the winners will be those who can embrace these changes effectively.