The Babylonian scholars who placed one season's astronomical observations beside another's were not being slow by choice.
They were being slow by necessity. The kind of intelligence they were building — the comparative kind, the kind that sees pattern across time rather than data at a point — cannot be produced quickly. It requires the earlier records to exist, to have been maintained in comparable form, and to be held in deliberate contact with the current record by someone with the knowledge to notice what the comparison reveals.
This is what comparative intelligence looks like at its operational level. Not a capability that a system provides, but a practice that a function performs. And the practice requires something that most organisational incentive structures actively work against: deliberate friction.
Not inefficiency. Friction has a specific meaning here. The structured resistance that forces a decision into contact with its own history before it is made. The explicit pause — built into the process, not left to individual initiative — that asks whether this situation resembles any situation the organisation has previously navigated, and if so, what happened.
In practice, this means building certain things that most organisations do not have.
It means keeping records of reasoning, not just outcomes. The decision reached is in the minutes. The reasoning behind it is often not. When the same decision point arises again, under slightly different conditions, the organisation that can access the previous reasoning is in a different position from the one that can only access the previous conclusion.
It means maintaining comparability across time. Long-term comparison is made structurally difficult when organisations change their information architecture — every system migration, every shift in reporting conventions, every redefinition of categories creates a break in the comparable record. Treating these as purely technical changes is how organisations quietly lose the ability to learn from their own history.
It means protecting the time of people who do the slow work. The analyst doing comparative work across years of records is doing something more important than the analyst running this week's report, but the second is more visible, more urgent-feeling, and more likely to attract institutional reward. The first is more likely to get absorbed by the second, which is how the slow intelligence disappears.
The discipline is not making all decisions slowly. It is knowing which decisions require the kind of intelligence that cannot be retrieved because it has not yet been produced — the kind that only emerges when someone sits with the record long enough to understand what the record, in combination with other records, is actually saying.
That requires friction. Build it deliberately, or the incentive structure will remove it quietly, and call the removal efficiency.