Pietro Schirano demonstrates an agent-connected design canvas and argues that software should accommodate users who bring their own agents and context. David Ondrej asks how people can supervise agent work and how products may adapt to those workflows.
Pietro Schirano describes short demonstrations, user feedback and community as ways to test and explain a product. His predictions about agents and software are presented as his views rather than established outcomes.
Pietro Schirano shows reusable text expansions for explaining code, planning work, reviewing changes and coordinating agents. The discussion closes with examples of using a coding agent to experiment with hardwareAn AI coding agent is a tool-using AI system that can inspect, modify, and validate software within a repository. and the importance of checking code for errors.
Watch on YouTube
