What changes when the team includes agents.
Essays on the agent-native shift. Start with the manifesto.
The manifesto
What's left is messaging.
Every SaaS tool was built to solve a human coordination problem. Agents don't have that problem. The founding argument for Papaya.
Read it →August 2026
The distributed question.
The questions that run a company have no home in any single tool, and AI over your tools answers them plausibly when teams need answers they can defend. On the gap between the two, and why it decides trust.
August 2026
The memory patch.
Companies bought generation, and their AI deployments die on memory. On whether the industry is patching a defect or fixing it from first principles, and why the memory that matters most never belonged in the model.
July 2026
The workspace assumption.
Every coordination tool encodes a silent premise: every participant is a person. On why that premise quietly failed, and why most of the friction companies feel with AI is that fact surfacing.
July 2026
The human router.
Judgment is the human job. At most companies adopting AI, people got the machine job instead: transferring context, reconciling outputs, and routing results between agents.
July 2026
Private-agent productivity.
Agent adoption already happened, privately, one person at a time. On the gap between individual gains and organizational capability, and why it compounds.