5 Comments
User's avatar
Jelle van Wieringen's avatar

I think it is wise when we start defining the terms, to work upwards from the building blocks, rather than working downwards from the use cases. The foundation is the AI agent, and I am seeking a robust definition of this concept for my role in moderating a hands-on course on agentic AI. Once this has been agreed upon, ANO or NAO will follow naturally. The most important question is whether AI agents are created on the fly or if humans stay in the loop. I have a gut feeling that we are not ready for that yet, as the potential impact is out of our imagination.

Expand full comment
Joseph de Castelnau's avatar

From your dry humour, I am gathering that you are grasping that maybe those agents are doing work that should not be there if we built better systems (more open, more composable ...)? Was that your message? How are you moving that ball forward? I stop investing in organizations that constantly gate knowledge or UX dark patterns (terrible way to start a relationship).

Expand full comment
Jurgen Appelo's avatar

Good point!

When we move from task-oriented processes to outcome-oriented agents, maybe there is a better chance we will get rid of pointless bureaucratic nonsense.

Expand full comment
Joseph de Castelnau's avatar

It does beyond bureaucratic nonsense. The economic system favors walled gardens at scale (closed platforms that are opened only in appearance). I am saying that a lot of the AI hype is about boilerplating those fixes (most of the one you describe about your travel experiences). How do we change the objective functions, reward system for the product we use and/or create?

Expand full comment