There was a question asked at the end of the Bonus Day video re: how to build your 'Who I Am' layer if you have multiple roles. I had the same question but I was also wondering about how to build context for multiple personas e.g., your product has several customer personas. If I want to create an AI agent for a customer persona to assist me with creating content targeted to that customer group. Is there a way to 'compartmentalize' context within one Claude/ChatGPT account?

What guidelines should we be following to protect our intellectual property when using publicly available AI models? Are there any references/resources that describe this? What happens to the content that we upload?

A lot of AI conversation still centers on output. Can it write the draft, summarize the meeting, build the outline, generate the ideas? Those are useful questions, but they can pull attention toward the most visible part of the workflow while hiding one of the biggest time drains underneath it. In many teams, the real issue is not that people cannot create. It is that they keep having to remember. Projects slow down because memory is scattered. Decisions live in old notes, in message threads, in slide comments, in someone’s head, and in documents no one has opened in two weeks. Then the team returns to the work and spends precious time reconstructing what already happened before anything new can move forward. That is why one of the most important shifts in AI right now is not just better generation. It is the emergence of AI as a project memory. ------------- Context ------------- Most projects do not fail because people stop caring. They slow down because continuity gets lost. A meeting happens, a decision gets made, a direction changes, and then the next piece of work begins without the full thread intact. Someone asks, “Did we already decide this?” Another person says, “I think that was in the notes somewhere.” Ten minutes later, the team is still trying to recover the state of the project before the real conversation can begin. This is not a small inconvenience. It is a structural time leak. It stretches cycle time, increases context switching, and quietly raises the cognitive cost of every task. Work becomes heavier because people are not only doing the work. They are also rebuilding the memory needed to do it well. That is why organized project memory matters so much. If AI can help preserve the running logic of a project, not just the latest output, then teams spend less time restarting and more time progressing. That changes the pace of work in a very practical way. The point is not that AI should remember everything. The point is that it can help preserve what matters most, decisions made, constraints set, priorities chosen, open questions, and the next useful step. When that memory is easier to access, the whole workflow becomes lighter.