Google scientist shares the 15-25-10 minute structure for job talks in academia and industry
docmilanfar · x · 2026-10-31
David Milanfar, deep learning scientist at Google, shared a recommended structure for job talks for those applying to academia or industry:
- First 15 minutes: an overview of your overall technical accomplishments, briefly covering a few representative works/papers — ideally tied together by a common theme.
- Next 20-25 minutes: a deep dive on just one of them — the one you're proudest of, or the one that best matches the team/department's interests.
- Final 5-10 minutes: summarize and discuss important new research directions you believe in, and their potential impact on the field, company, or department.
His verdict: it's hard to give a bad talk structured this way.
More from Companies & People
- Toyota and Waymo to launch bZ4X robotaxi pilot in Japan in 2027 — Scobleizer · 2026-10-31
- Google AI Overviews now intercepting even navigational searches, founders complain — bindureddy · 2026-10-31
- Factory ditches .ai, snags factory.com to pitch agent-native software development — matanSF · 2026-10-31
- Baidu's Apollo Go rides into London and launches Hong Kong's largest public robotaxi trial — Baidu_Inc · 2026-10-31
- Google AI Ultra signup is a mess: upgrade button just emails your Workspace admin — npinto · 2026-10-31
- Wispr Flow drops resumes after 2,000 identical applications, hires by voice-built projects — Arindam_1729 · 2026-10-31