Open notebooklm.google.com and start a new notebook. Paste or upload a short public page about AI agents (or use the quote from SNK3-065). Ask: 'What risks does this source mention?' You're done when the answer cites your source — or honestly admits it found no risk text.
How-to · INF · INF.1
X-style: NotebookLM meets the feed
Bring feed ideas into NotebookLM (or grounded tools) so answers cite sources, not hype.
Worked example
Combining a Tim-priority tool with feed culture: upload a public article about agents into NotebookLM and ask it grounded questions — instead of letting a chat freely hallucinate a summary.