Skip to main content
The Human Bit
← All capabilities

Capability guide · Plain language first

Work with files and data

Pull information out of documents, tables and spreadsheets, then work out what is actually going on inside them.

The pattern

See what changes, and where you return.

Files or a table go through analysis, then a person verifies the finding against the original.

Recognisable examples

Begin with work you already understand.

A capability is a broad kind of help. The linked task guides explain how to complete a specific piece of work.

01

Make sense of a spreadsheet someone handed you

02

Find the recurring themes in a pile of customer feedback

03

Compare two contracts for what genuinely differs

04

Draft a chart from figures you provide

AI can contribute

Repeatable effort and a first pass

  • Reads files you upload and describes how they are put together
  • Runs the analysis in a sandbox and can show its working
  • Finds patterns, outliers and likely data-quality problems
  • Can read a chart or photo you attach and pull the figures out

The person contributes

Meaning, checking and responsibility

  • Explains what the columns and documents actually mean
  • Checks the numbers back against the original source
  • Decides whether a finding is real or just a coincidence

Featured Human Partnership

Research becomes useful when it changes the quality of a decision. A long report is not evidence of good research unless the scope, sources, disagreements and remaining uncertainty are visible.

AI capability
Editable research planning. Deep research features can turn a broad question into a proposed plan, source strategy and sequence of subquestions that the person can review before a long search begins.
Human expertise
Decision framing. You know what action the research must support, which uncertainty is blocking progress, and what level of confidence is sufficient for the consequence involved.
Decision
The person decides whether the evidence is credible, relevant and sufficient for the real decision. AI can assemble and challenge a case, but it cannot set the organisation's evidence threshold or accept the consequence of being wrong.
Skill built
Evidence judgement. Used well, the workflow strengthens question framing, source evaluation, uncertainty handling and the discipline of matching the strength of a conclusion to the strength of its evidence.
Finished result
A focused briefing that keeps its links, its disagreements, its gaps and your next questions all in view.
Open the complete Human Partnership workflow →
  1. 01Turn the topic into a decision

    State what must be decided, what is already known, the time and geographic boundaries, and which source types should carry the most weight.

  2. 02Propose the research route

    Generate a research plan, candidate questions and source strategy for review before using deep research, web search, uploaded files or connected sources.

  3. 03Interrogate the emerging evidence

    Refine the search when terms are ambiguous, important viewpoints are absent, or early evidence suggests the original framing was too narrow.

  4. 04Decide what the evidence earns

    The person decides which conclusions are supported, which remain provisional, and whether more evidence, a subject matter expert or a real world test is required.

  5. 05Open the claims that carry the decision

    Read the most important sources directly, check publication dates and jurisdiction, trace quotations and numbers, and record what the search could not find.

Continue from here

Related tasks and lessons

Reviewed
Review by
Owner
The Human Bit editorial