When work fragments between humans and AI: the continuity challenge

We have never had so much context available, and yet it is getting harder and harder to know where we stand.

As LLMs settle into our work, our thinking fragments across conversations, tools and models. The phenomenon intensifies further with agents, to which we sometimes delegate a large part of the work, and even decisions.

We work a great deal on how to give context to models, and far less on how to preserve the continuity of the work between humans and AI: what has been decided, explored, abandoned, delegated, or is still open.

This keynote explores that shift of problem: as work is distributed across humans, LLMs and agents, how do we keep hold of its thread?