Purpose
Use this lesson to use AI to organize research while separating sources from claims. The aim is to produce work that a colleague can review and act on, not merely text that sounds convincing.
Lesson Explanation
Scenario: Research whether a market segment is growing.
Better instruction: Create a research brief with claims, supporting source links, publication dates, and questions that remain unanswered. Do not cite sources you cannot identify.
Research output becomes useful when every important claim can be traced. Request a claim-and-source structure, date limits, and an uncertainty section. The model may help plan search terms or summarize supplied sources, but source checking remains a human task. Each major claim has an identifiable source and date, and unknowns are listed.
Work in a short cycle: provide the relevant input, state the result and limits, inspect the draft against the acceptance criteria, then revise the instruction when a requirement is missing. Keep the final human decision with the person responsible for the work.
Practice Questions
-
A prompt asks an AI to “research our competitor’s pricing” with no requirement to cite sources. What does this lesson say research output needs to become useful?
View Answer
Every important claim needs to be traceable to a source – a claim-and-source structure, date limits, and an uncertainty section.
-
What does this lesson say remains a human task, even when the AI helps plan or summarize research?
View Answer
Source checking.
-
A research brief includes a confident claim with no citation attached. What does this lesson say should happen with this claim?
View Answer
It should not be presented as verified – every important claim should be traceable to a source, and unsupported claims should be flagged as such.
-
Write a prompt instruction for a research brief that follows this lesson’s claim-and-source structure.
View Answer
Something like: “For each claim, provide the supporting source link and publication date. List any questions that remain unanswered. Do not cite a source you cannot identify.”
-
A research brief cites a source, but the source’s publication date is more than five years old for a fast-moving topic. What element of this lesson’s guidance addresses this kind of issue?
View Answer
Date limits, part of the claim-and-source structure this lesson recommends specifying.
-
Why does this lesson explicitly say the model “may help plan search terms or summarize supplied sources,” but not replace source checking?
View Answer
Because verifying that a source is real, current, and actually supports the claim is described as a human task that shouldn’t be delegated entirely to the AI.
-
A completed research brief includes a section listing questions that could not be answered with available sources. What element of this lesson’s recommended structure does this reflect?
View Answer
An uncertainty section.
-
A prompt for a research brief doesn’t instruct the AI on what to do if it can’t identify a real source for a claim. What specific guidance from this lesson addresses this gap?
View Answer
Do not cite sources you cannot identify – an explicit instruction to avoid inventing citations.
-
A market research brief has every major claim linked to a specific, dated source, with an unknowns section listing open questions. What does this reflect?
View Answer
A properly structured research brief, following all the elements this lesson recommends: claim-and-source structure, dates, and an uncertainty section.
-
Why might a research brief that “sounds well-researched” still be unreliable, based on this lesson’s core concern?
View Answer
Confident, well-written prose doesn’t guarantee that claims are actually traceable to real, verifiable sources – the structure and traceability matter more than the writing quality.
-
A prompt asks for research on an emerging market trend, specifying that all sources must be from the last two years. What element of this lesson’s guidance does this reflect?
View Answer
Date limits.
-
A colleague receives an AI-generated research brief and immediately shares it externally without opening any of the cited sources first. What does this lesson suggest about this practice?
View Answer
Since source checking remains a human task, the sources should be verified before the brief is used externally, not simply trusted because they are cited.