Every organisation is using tools to think together. Surveys, workshops, strategy off-sites, AI-generated summaries. These tools are not poorly designed. Many of them are excellently designed for their specific purpose. The problem is that their specific purpose is not what organisations actually need when they face complex, consequential decisions.

What surveys destroy

A survey achieves statistical representation by isolating responses: each respondent answers independently so responses can be aggregated without the contamination of social influence. This is methodologically sound. But the isolation that makes surveys statistically valid also destroys what is most valuable about human understanding: the way perspectives develop in relation to each other.

A survey tells you what people think. It cannot tell you from where they are thinking — or what they would understand if they heard each other.

What workshops destroy

A workshop achieves in-room alignment through facilitation that moves a group toward a shared conclusion within bounded time. But the convergence a workshop produces is shaped by three forces unrelated to the quality of collective understanding: the status of the most senior person in the room, the confidence of the most vocal contributor, and the time pressure of the format. Workshops also expire. The understanding built in a two-day off-site begins to degrade the moment the flip chart is rolled up.

What AI destroys

AI is optimised for synthesis of existing expressed content at unprecedented scale and speed. But synthesis of existing expressed content is the most conservative possible approach to complex questions. The understanding that hasn't yet been expressed — the tacit knowledge that exists in practice but not in documents, the perspectives of those who were not asked — are invisible to AI.

The common failure mode

All three tools share a structural failure: they are designed to aggregate or synthesise existing expressed positions. They produce outputs that look like understanding — a survey report, a workshop summary, an AI brief — but have discarded the texture, the reasoning, the epistemic ground, and the relational development that genuine understanding requires. The result is decisions made from an impoverished picture.

What the alternative requires

Genuine collective understanding requires: perspectives that carry their epistemic ground; a structure that enables perspectives to develop in relation to each other; a format that does not expire or converge prematurely; and a way to surface what the collective understands without flattening the texture of how it got there. That is a different category of tool. A collective sensemaking operating system.