The article examines the current state of AI adoption, particularly in vertical markets, arguing that the conversation often swings between overly optimistic and excessively cautious narratives. Instead of seeing AI as a binary replacement for human labor or a series of failed pilots, the author highlights a more complex reality. The focus is on the diffusion of large language models (LLMs) and how they integrate into specific industries. The piece introduces two frameworks: Rogers’ “Diffusion of Innovations,” which outlines how technologies spread, and Evans’ “Absorb/Innovate/Disrupt,” which describes the stages of technology adoption. Rogers’ model identifies key factors influencing adoption, such as relative advantage and trialability. Generalized LLMs score well on these but face challenges in compatibility and complexity when applied to vertical markets. For example, while marketing teams might easily adopt AI for content creation, construction managers struggle with integrating AI into their established workflows. The article notes that most vertical markets are still in the early Innovators phase of adoption and emphasizes that bridging the gap to the early majority requires more than just improving technology; it necessitates aligning AI tools with existing work processes. The Absorb phase of Evans’ framework is particularly relevant, where organizations begin to automate existing workflows to improve efficiency. Successful applications of generative AI today—like coding assistants and customer support automation—showcase this stage by enhancing established practices. The article identifies two main areas for current AI applications: documentation and customer communication. Despite these early successes, significant barriers still exist, particularly concerning human capital and integration with legacy systems. Human expertise in both technology and the specific industry is crucial for successful implementation, and integration challenges arise from the fragmented nature of existing systems.