Skip to content
Silicon Valley GirlSilicon Valley Girl

Head of Gemini: You're Using 5% of What Gemini Can Actually Do | Josh Woodward

📌 Increase your visibility in AI search with HubSpot AEO — see how your brand shows up in ChatGPT, Gemini, and Perplexity. FREE for 28 days, then $50/month: https://clickhubspot.com/a400ee Josh Woodward, the VP behind Google Labs, the Gemini app, and AI Studio, gave me the clearest picture yet of where AI is actually heading at Google. 0:00 — Intro 1:01 — Why This Google I/O Is Different From Every Other One 2:21 — Feature That Was Built Two Weekends Before I/O 3:22 — Why Switch to Gemini When You Already Use Another AI 6:53 — The Shift to Voice-First Is Already Happening 8:02 — From Doing to Directing: Everyone Becomes a Manager 12:10 — Is Google Actually Losing the AI Race? 15:18 — Gemini's Personality 16:27 — How Josh Builds Personal Context Into AI 19:58 — Why Everyone Needs to Organize Their Personal Context Now 20:55 — What AGI Actually Means 22:35 — Why Human Taste Becomes More Valuable, Not Less 23:42 — Advice for Someone Starting Their Career Right Now 24:29 — How Google Labs Ships Products in Two Weekends 26:13 — The Metric Josh Uses Instead of Retention Dashboards 29:39 — The Next Shift Nobody Has Internalized Yet 31:12 — What His Five-Year-Old's Future Looks Like *Links*: 📩 Follow my Newsletter: https://siliconvalleygirl.beehiiv.com/subscribe?utm_source=youtube&utm_medium=video&utm_campaign=futureproof-sub&utm_content=Josh Woodward 🔗 My Instagram: https://www.instagram.com/siliconvalleygirl/ 📌 My Companies & Products: https://Marinamogilko.co #googleio #podcast #gemini #geminispark

Josh WoodwardguestMarina Mogilkohost
May 21, 202633mWatch on YouTube ↗

At a glance

WHAT IT’S REALLY ABOUT

Gemini’s agent era: voice-first assistants, personal context, massive speed gains

  1. Google I/O marked a shift toward an “agentic era,” where Gemini Spark and related tools run tasks in the background across Gmail, Drive, Calendar, and more.
  2. Gemini’s differentiators are tight Google ecosystem integration, scalable parallel task execution via cloud VMs, and a broad generative media stack (docs, slides, images, video, music).
  3. Voice-first interaction is already tipping in some countries, and new features like Docs Live and voice-to-workflow demos show voice moving beyond transcription into tool-calling and synthesis.
  4. Woodward argues the real user shift is from “doing” to “directing,” meaning people will increasingly manage agents and need new intuition, soft skills, and judgment rather than only execution skills.
  5. In Google Labs, rapid iteration with small teams and real-world user testing (the “eyes light up” metric) is emphasized over early dashboard optimization, because most ideas take multiple attempts before success.

IDEAS WORTH REMEMBERING

5 ideas

Gemini’s pitch is “no connectors” for Google users.

Woodward frames Gemini Spark’s advantage as native access to Gmail, Drive, Docs, Sheets, Slides, and Calendar via an opt-in “Personal Intelligence” concept, reducing setup friction versus external tools.

Agents will scale from a few tasks to hundreds running in parallel.

Gemini can spin up cloud virtual machines to execute many background tasks simultaneously, positioning agent workflows as an orchestration problem rather than a single-chat interaction.

Voice is becoming a primary interface, not a novelty.

He notes usage has tipped toward voice in some countries, and upcoming features combine voice input with tool calling, retrieval from files, and automatic cleanup of rambling speech into polished outputs.

A killer demo is “talk to your files, then ship a deliverable.”

The two-week-old I/O demo showed selecting multiple files (Drive/desktop), speaking instructions, and having Gemini synthesize PDFs/images into a structured email or document—correcting errors like dates along the way.

The work shift is from execution to management of outcomes.

Woodward describes moving from “doing” to “directing,” implying broad “manager training for everyone” as individuals coordinate multiple agents toward a final deliverable.

WORDS WORTH SAVING

5 quotes

And so we on the team talk about is you're moving from doing to directing, and that's, like, a big shift.

— Josh Woodward

Now we're imagining we may need that for everybody... because you may be managing these different sort of, like, agents and others, so yeah.

— Josh Woodward

I always kind of tell the teams, and I try to go with them on a lot of these, you see it in people's eyes. When they use the thing, do their eyes light up, or are they, like, recoil?

— Josh Woodward

The other one I am not sure people have fully internalized, even at Google, is how fast these models are gonna get.

— Josh Woodward

So speed's a feature.

— Josh Woodward

Gemini Spark and “Personal Intelligence” opt-in contextDeep integration with Gmail/Drive/Docs/CalendarVoice-first UX and Docs LiveAgent orchestration: “doing to directing”Model speed as a product feature (tokens/sec)Personal context management (notebooks, notes, best writing)Google Labs iteration culture and user-testing heuristics

High quality AI-generated summary created from speaker-labeled transcript.

Get more out of YouTube videos.

High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.

iOSAndroidClaudeChrome