‹ All demos
Panel Prep Kit · For the February Panel

A research workspace that only answers from the evidence

Fifteen minutes of setup gives you a private research assistant for the panel, built on Google NotebookLM. It answers only from the six sources below — every reply carries a citation back to the exact passage, and when the answer isn't in the sources, it says so instead of guessing.

Setup: ~15 minutesCost: freeRequires: any Google account
1

Create the notebook

Go to notebooklm.google.com, sign in, and click “+ Create new notebook.” Name it Museums & AI — ASMD Panel Prep.

2

Add the six sources

For each, click “+ Add source” → “Website,” paste the URL, and let it ingest. The order runs from sector frameworks → ethics → policy precedent → current state.

SourceWhy it's in the kit
iASTC — “Building Public Agency in AI”The science-center sector's own framework: 13 roles museums can play in public AI engagement.astc.org/ai
iiSmithsonian — AI Values StatementThe most-cited values document in the field; a living document since 2022.datascience.si.edu
iiiSmithsonian — AI Usage StatementThe practical companion — how to disclose AI use: content format + role of staff + role of AI.si.edu/about
ivAAM — “Beyond the IT Office”Argues AI policy belongs at leadership level, not IT — and carries the survey data.aam-us.org, June 2025
vAAM — “AI in Museums: Ethics & Protocols”Concrete protocol-building steps and example policies from the 2025 AAM session.aam-us.org, July 2025
viASMD — 2025 Directors Forum recapAI was already on the Forum agenda — useful for framing “where we left off.”asmd-us.org
If a page imports as a thin stub: open it, copy the article body, and use “Paste text” instead. NotebookLM does better with full text than a sparse crawl.
3

Generate the overview

With all six loaded, open the Notebook guide and generate the summary — a two-minute orientation to what the corpus covers. The optional Audio Overview turns the sources into a ~10-minute podcast-style discussion, good for the drive.

4

Five questions worth asking it

Paste any of these into the chat. Every answer cites which source it came from — click the citation to verify before you quote it on stage.

Q1 · Talking points
Across all sources, summarize the 5 most important things a science-museum director needs to understand about AI right now. Keep each to two sentences and cite the source.
Q2 · Objection prep
If a skeptical board member asks “Why does our museum need an AI policy at all?”, what evidence-backed answer do these sources support? Include the strongest counterpoints I should be ready for.
Q3 · Citation finder
What specific statistics or survey numbers about museum AI adoption appear in these sources? Give the number, what it measures, the sample size, and the exact source for each.
Q4 · Framework comparison
Compare how the ASTC typology, the Smithsonian statements, and the AAM materials each handle disclosure and transparency of AI-generated content. Where do they agree, and where do they differ?
Q5 · The policy gap
These sources suggest most museums don't yet have an AI policy. Based only on the precedents here (e.g. the Smithsonian), what would a minimal, responsible first-draft AI policy contain for a mid-sized science museum?

Three assurances

⭳ Download this guide as markdown