The number changed. Did the story?
Why a research record needs the first release as well as the latest value.
Research
Analysis that separates the observation from the conclusion.
Browse research
21 results
Enter a title, subject, author or institution.
Before the percentage, find the denominator
A total becomes informative when the population, stage and time window are explicit.
The evidence between a promise and an outcome
An opening research agenda on capacity, recognition and the conditions that make a claim usable.
485 TWh to 950 TWh: an outlook, not a delivery ledger
The IEA’s 2026 data-centre projection needs both its baseline and its uncertainty label.
A generation queue is not a data-centre queue
Berkeley Lab’s 2026 edition is a valuable dataset with a specific unit of observation.
A policy rate is not a project’s cost of capital
The rate decision is one input. The financing contract carries the conditions.
The skill remains. The door can still close.
Separate practiced ability, institutional recognition and the procedure that connects a person to an opportunity.
The AI revenue ladder has missing rungs
A worked Microsoft example separates investment, lease recognition, cloud revenue and the missing evidence of asset utilization.
From a rule on paper to an evidence trail
Compliance claims become researchable when the obligation, actor and observable record are connected.
Follow the evidence
Use the Knowledge library to find original papers and documents cited across the publication.