Designing Your Co-Intelligence
A personal system of AI collaborators, knowledge infrastructure and working practices designed to extend what a person can think and do.
Useful co-intelligence starts with self-knowledge. Your strengths, blind spots, working preferences and sources of friction become design requirements for support that complements how you actually think.
The question
What if we stopped measuring AI augmentation primarily by how much work it removes, and started examining what it makes possible for people to think and do?
Saved time is useful, but it captures only one part of augmentation. I am interested in how AI can extend judgement, preserve context, introduce productive disagreement and supply capabilities one person cannot bring equally strongly to every task.
Most AI advice starts with tools and tasks. My work begins with the person: what they already do well, where their thinking benefits from challenge or structure, what needs to be remembered and how different forms of support can work together.
The design approach
I treat co-intelligence as an environment shaped by four connected design problems.
-
01 · Self-knowledge
Know the intelligence you are extending
Map strengths, blind spots, preferences, friction and the conditions that improve judgement.
-
02 · Complementarity
Choose complementary capabilities
Design collaborators with distinct remits, positions and standards. Productive incompatibility broadens the range of available thinking.
-
03 · Continuity
Give collaboration memory
Preserve sources, context, decisions and developing ideas so useful work accumulates across conversations.
-
04 · Working system
Fit the design around real cognition
Shape tools, handoffs and boundaries around how the person works, then evaluate what becomes easier, stronger or newly possible.
Work in practice
The ideas are developing through systems I use. These are the clearest examples so far.
-
Self-knowledge system
Inner Magpie
A private, longitudinal record of what I learn about how I think, work and live.
-
Complementary team
Designing AI collaborators
A deliberately assembled team with different remits, perspectives and standards.
-
Knowledge infrastructure
Designing knowledge and workflows
Systems that capture, classify and recover useful thinking without slowing it to the speed of manual knowledge management.
This page will collect further notes, frameworks, tools and experiments as the work develops.
The book
Design Your Co-Intelligence develops the foundation of this work: before you can deliberately design a productive relationship with artificial intelligence, you need a clearer model of your own. The book brings behavioural science, metacognition and interaction design together for people whose work depends on complex judgement.
Workshops and training
I am developing practical workshop formats for people and organisations that want to design useful co-intelligence around real work. Possible areas include:
- mapping the human capabilities and friction a system needs to complement
- defining roles, remits and boundaries for AI collaborators
- designing memory and knowledge workflows around collaboration
- evaluating augmentation through what becomes newly possible
If you are exploring a workshop, training session or collaboration in this area, get in touch.