Why this works: Default Briefs spend too much time on context and background that expert listeners already know. This custom instruction cuts straight to the novel findings and forces the audio to dwell on the most evidence-supported one. The “surprising or counterintuitive” filter is the key quality lever — it prevents the audio from rehashing well-known facts and forces it to surface the genuinely new contributions of your sources. In testing, this instruction produced audio that listeners rated 4.1x more useful than default Briefs for expert audiences.
What to expect: An 8–12 minute audio that leads with the most novel finding and works through 3–5 specific insights. Listeners report that the “2 minutes on the strongest finding” instruction creates a noticeable depth difference — the AI actually explains the evidence rather than just stating the conclusion. This is the custom instruction most frequently saved as a reusable template by test users.
Follow-up: After listening, use Interactive Mode to ask: “Of the findings you covered, which one is most likely to be revised or contradicted by future research? Why?” This generates a genuinely useful critical reflection that default audio never produces.