I Don't Even Know What to Build Anymore (vancelucas.com)

A growing sentiment among developers highlights a significant shift in software design, emphasizing the need to create applications tailored for AI agents rather than traditional user interfaces. As reliance on data-driven systems increases, developers are grappling with the implications of building tools that facilitate agentic workflows through APIs, CLIs, and MCPs, as opposed to the conventional dashboards and visualizations. This shift raises questions about the future value of products, the nature of data ownership, and the risk of competitors (including major AI labs) surpassing individual creations with their own integrated solutions. The uncertainty surrounding the viability of software ideas in an AI-dominated landscape is palpable, leading developers to rethink existing concepts that previously thrived in a more static environment. While AI accelerates the development process, it also makes it challenging to identify sustainable business models, pushing many back to the drawing board. This reflection underscores a broader anxiety within the AI/ML community about navigating an evolving market where the successful paths are not yet clear, prompting a reevaluation of strategies and aspirations as the industry experiences a fundamental paradigm shift.