AI Research Summary Template With Sources, Assumptions, and Next Steps

# AI Research Source Pack

## Research question
What were we trying to learn?

## Short answer
Give the current answer in 3-5 sentences.

## Source table
| Source | What it supports | Confidence | Notes |
| --- | --- | --- | --- |
|  |  | High / Medium / Low |  |

## Key findings
1.
2.
3.

## Assumptions and limits
- What might be outdated?
- What did the AI not verify?
- What sources disagree?
- What would change the answer?

## Decision or recommendation
What should the team do with this research?

## Next action
- Owner:
- Due:
- Smallest next step:

## Reusable context
What should be saved for future AI chats, agents, or project briefs?

- Google blog
- NotebookLM help
- Reddit thread

| Google Gemini notebook announcement | Supports the claim that Gemini notebooks sync with NotebookLM | High | Official product announcement |
| NotebookLM public notebook help | Supports public sharing behavior and chat view caveat | High | Official help page |
| Reddit discussion | Shows user concerns and language, not product truth | Low | Use as intent signal only |

| Official doc | Supports platform behavior or product limits | High | Link to the exact help/doc page |
| Community post | Shows how users describe the problem | Low | Intent signal only, not proof |
| Internal note | Shows team decision or prior context | Medium | Remove private details before sharing |
| AI transcript | Shows reasoning path and selected turns | Medium | Include only cleaned, relevant turns |

# AI Research Source Pack

## Research question
Should we send a teammate the whole NotebookLM notebook or a cleaned research handoff?

## Short answer
Use the public notebook when the teammate needs to explore the full source-backed workspace. Use a cleaned handoff when they need the answer, supporting sources, caveats, and next action. Chat View can narrow the default experience, but it should not be treated as a guarantee that underlying notebook contents are fully hidden.

## Source table
| Source | What it supports | Confidence | Notes |
| --- | --- | --- | --- |
| Google NotebookLM public notebooks help | Public sharing behavior, Google account requirement, Chat View caveat | High | Official help article |
| Google Gemini notebooks announcement | Gemini notebooks can organize chats, files, and source-backed project work | High | Official product announcement |
| Team's cleaned AI chat | Shows the actual question, tradeoffs, and decision language | Medium | Internal evidence; remove private context before sharing |

## Key findings
1. A public notebook is best when exploration matters.
2. A research handoff is better when the reader needs to act quickly.
3. Source labels and caveats should travel with the answer.

## Assumptions and limits
- Product sharing behavior may change.
- The teammate may not have the same account or access context.
- The AI-generated answer still needs source review.

## Decision or recommendation
Send a cleaned research handoff first. Include the public notebook only as an optional reference if the teammate needs deeper exploration.

## Next action
- Owner: research lead
- Due: before the team sync
- Smallest next step: convert the useful AI turns into a source-backed handoff link

Research handoff:
- Question:
- Short answer:
- Best source:
- Biggest caveat:
- Decision needed:
- Next action:
- Source pack: