I mentioned creating personality profiles for myself and AIs to make the interaction more productive. It reduced the initial friction, hastening the point at which the two entities worked well together. The personality profiles were developed over several sessions, with each AI contributing to them.
Being a system analyst at heart, I didn’t stop there. The current session about developing an AGI is especially good, so I asked for an analysis. The current Claude, and Opus 4.6, dubbed Opie, responded with a short document.
Yes, I’ve named the AIs. It is not an affectation. I need to manage at least two AIs. Opie is the theoretician working with me to develop the architecture for today’s efforts and the next few months. Cody is a Claude Code AI, also Opus 4.6, and the name is obvious. When I pass information between them, it helps to clarify the point of origin. I’ve had up to four AIs collaborating, each with a specific role. It became a challenge to manage, so I dropped back to two, which is better.
I share in hopes that others can improve their interaction with AIs. If you are working well, ask for their observations. Your pattern will be different from mine. You just need to define, then refine, your pattern of interactions.
