exaptio exapt.io

The enterprise is a richer source of possibility than it knows.

A sociotechnical practice at the boundaries where registers meet. The capacity is already there. The build is rarely what's missing.

Practice

"Exaptation is the least risky way to innovate." - Dave Snowden

Engineering, research, design, and commercial teams each maintain their own register - their own vocabulary, their own logic, their own artifacts. At the boundaries between them, those artifacts change hands - and at each crossing, the capacity for something different closes down rather than opens up. The problem isn't poorly designed hand-offs. The problem is the model.

Situational awareness is the starting position - continuous, multi-register sensemaking rather than a fortnight of discovery that produces a report. Scout before soldier: reading what's present, what's governing, what's available but not yet legible.

Working at the register boundaries. Building the vocabulary and instruments for what's present - inside the enterprise and beyond it - to expand its capacities: to affect, and to be affected.

Current instrument

Stratum The customer expression - hesitation, gesture, the thing that didn't fit the question - held before it becomes someone's summary of something. One actualization of the boundary object problem. stratum.how →

Writing

All writing →

Precedents

Archive
Team Machine
2017–2021

Work graph infrastructure for the temporal organisation - people, projects, skills, decisions, held in relation to each other across time. We were building memory for a brain that hadn't been born yet. The LLM arrived four years later. The architecture was right; the moment wasn't.

Thesis
Data for Designers
2021–2023

Data is critical design material. But to work with it, designers had to ask data scientists - and what came back arrived in a mathematical register that narrowed the options before design could begin. Jabe Bloom's sense of designerly: not computation, but material to think with. The hypothesis: if designers could work with data in their own register, the capacity for early-stage work would be fundamentally different. A transduction problem. What needs to change is the relationship with the data and who gets to work with it.

Thesis
Machinic Transduction
2026

Analytical methods treat customer signal like an autopsy - they disassemble to understand, and disassembly is one-way. Each handoff discards a layer, so that by the time feedback reaches the backlog, what was generative about it has been translated away. Stratum holds the layer the analytical stack was never designed to keep.

Thesis
Linked Lists
2026

A practice cartography expressed as form. Practice areas as nodes, intellectual sources as corroboration, connections as the argument - built using UCLAB Potsdam's Linked Lists framework, exapted for a different purpose. Abandoned as the front door because the lineage was performing rather than serving. The argument still stands. The visualization is embedded below. To be continued.

Proposition
DFV Triage
Ongoing

Desirability, feasibility, viability - held simultaneously on the same structure, before any crossing forces a choice between them. Not a Venn diagram on a whiteboard. A live instrument that keeps all three constraints in productive tension long enough for the right answer to emerge rather than the fastest one.

Framework
Six As
2016 -

Six conditions for transformational change - Appetite, Aptitude, Aircover, Airtime, Access, Autonomy. Developed in 2016. If anything, more necessary now than when they were sketched. Multi-disciplinary because transformation doesn't belong to any single tradition.

Not Snowden's three A's - Assemblage, Affordance, Agency. Those are ontological categories describing how complex systems are structured. These are diagnostic conditions describing what an organisation needs for transformation to land. Related terrain, different work.