Lenny's PodcastHow to be more innovative | Sam Schillace (Microsoft deputy CTO, creator of Google Docs)
CHAPTERS
- 0:00 – 0:35
Work doesn’t have to be miserable: “Do the thing you feel guilty getting paid for”
Sam opens with a contrarian career philosophy: people undervalue what comes easily and assume valuable work must be unpleasant. He argues the highest-impact work often comes from leaning into what feels fun, even if it looks like “messing around.”
- •People discount skills that feel easy or enjoyable
- •The best work can come from playful exploration, not grinding
- •If you find something people will pay for that you enjoy, go all-in
- •Impact often correlates with doing what you’re naturally great at
- 0:35 – 3:53
Setting the stage: Sam’s resume and why innovation is the theme
Lenny introduces Sam’s background—Writely/Google Docs, multiple startups, leadership roles at Google, Box, and now Microsoft. The conversation focus is framed around disruptive innovation, original thinking, optimism, and career lessons.
- •Sam’s career spans startups, Google, Box, and Microsoft leadership
- •Writely became the foundation for Google Docs/Workspace
- •Core episode focus: how to be more innovative and think bigger
- •Innovation as a repeatable practice, not a one-off stroke of genius
- 3:53 – 6:46
The first Google Doc (and the “Document of Theseus” problem)
Sam describes the very first Google Doc file and why it still “works” despite multiple rewrites and migrations. He explains early infrastructure hacks and how preserving user data through major platform shifts became a quiet but meaningful accomplishment.
- •Writely began in 2005 in C# on rented Windows servers
- •Multiple rewrites and migrations (Java, Bigtable, Spanner; frontend rewrites)
- •Philosophical question: is it the same document after everything changed?
- •Early collaboration testing: collision detection, presence, pasted images
- 6:46 – 10:26
What makes innovation disruptive: “why not” vs “what if”
Sam explains why new ideas feel threatening and how people default to “why not” objections. He argues the innovators’ advantage is to ask “what if this works?” and follow the implications forward, even when skepticism is loud.
- •New ideas trigger identity/worldview threat, leading to rejection
- •“Why not” questions are easy; “what if” questions open possibility
- •Early Google Docs skepticism: trust, connectivity, airplanes
- •Engineers skew pessimistic; innovation benefits from chosen optimism
- 10:26 – 15:46
How to recognize disruptive ideas: toys, polarization, and strong reactions
Sam offers a detection heuristic: disruptive ideas often look dumb first and get labeled as “toys.” Another signal is bifurcation—people love it or want it to die—whereas incremental products produce mild indifference.
- •“Every new idea looked dumb at first” (and so do dumb ideas)
- •“It’s a toy” can be a sign it’s actually threatening/real
- •Love/hate response is a stronger disruption indicator than broad lukewarm feedback
- •Early Docs had intense boosters and intense detractors—even internally
- 15:46 – 19:47
Optimism as a practice: making experiments cheap and learning fast
Sam reframes optimism as a conscious operating mode that increases learning speed. By lowering the cost of experiments and staying receptive to surprising results, teams are more likely to discover breakthrough paths rather than quit early.
- •Optimism ≠ carelessness; it’s closer to growth mindset
- •You miss more opportunities by pessimism than you save by being “right”
- •Make experiments fast/cheap so you can try more things
- •Examples: multi-agent chat experiments; shared “whiteboard memory” made systems smarter
- 19:47 – 21:54
“Get to the edge and fuck around”: a method for invention
They discuss Sam’s blunt mantra for innovation: push to the frontier of what tools allow, then run many small experiments. Sam emphasizes attentiveness—breakthroughs often appear as interesting failures or unexpected behaviors.
- •Go to the edge of tools/technology; run lots of small trials
- •Being receptive matters as much as experimenting
- •Surprising failures can reveal the real product shape
- •Experimentation is a muscle—build it through repetition
- 21:54 – 28:25
User value beats hype: “People are lazy” and convenience wins
Sam explains his core product principle: adoption is driven by net user value minus the total effort required (learning, onboarding, habit change). He uses crypto as a counterexample—hard to find compelling user value even in the best case.
- •If you can’t articulate real user value, marketing won’t save you
- •Users consider the whole energy cost: discover, learn, remember, build habits
- •Convenience and reduced friction often dominate feature lists
- •Writely onboarding minimized friction (no email initially; gentle prompt later)
- 28:25 – 37:15
Building Writely/Google Docs: early trade-offs, collaboration pain, and internal resistance
Sam recounts how the team quickly felt the promise, then spent years solving collaboration and browser inconsistencies. He also describes how even after acquisition, major internal pushback at Google could have killed the project without protection and persistence.
- •Collaboration value was discovered early, partly “by accident”
- •Hard technical reality: browser DOM differences made merging gnarly
- •Early system used three-way merge before modern OT approaches
- •Internal opposition at Google was intense; survival required stubbornness + support
- 37:15 – 40:00
Product–market fit isn’t one moment: attention spikes, feature debates, and what mattered
Sam argues PMF depends on which market you mean; early energy appeared within months, but the long journey involved constant recalibration. He reflects on what was worth building (collaboration) versus costly detours (offline) and the later feature catch-up race.
- •Early traction and “curve” recognition drove attention from press/VCs
- •Docs was a second proof point after Gmail that real web apps were coming
- •Trade-offs vs Office: simpler UX + zero install + collaboration vs feature breadth
- •Offline work consumed effort but mattered less than expected
- 40:00 – 44:06
The future of documents: bots-as-docs, dynamic intent, and interactive artifacts
Sam predicts documents will evolve beyond static “wood pulp emulation” into conversational, semantic, interactive objects. He describes workflows where a bot interviews you to produce a readable artifact plus structured memory you can query, transform, and visualize.
- •“Bots are docs”: conversational artifacts become the new document primitive
- •Vector memory + dialogue enables querying, summarizing, and generating diagrams
- •Soon, non-interactive apps will feel “broken,” like offline devices today
- •Shift from static GUIs to intent-driven, adaptive experiences
- 44:06 – 49:22
When playing with tech is valid: pick “north stars” to focus experimentation
Sam defends tech exploration, but warns that unfocused tinkering rarely yields value. His approach is to define ambitious, testable “north star” goals (e.g., multi-agent autonomy), then learn by pushing toward them and building tooling along the way.
- •Exploration is good when anchored to a concrete target
- •North-star challenges reveal system gaps (debugging, monitoring, documentation)
- •Agents building tools for themselves (debugger agent, doc agent) can compound progress
- •Broad full-stack familiarity can unlock unusual combinations and insights
- 49:22 – 58:25
Thinking in the future requires risk tolerance: play, failure, and “virtue from error”
Sam connects future-oriented thinking to a willingness to take risks and look foolish. He argues modern incentives (curation, public failure, unrealistic success narratives) reduce experimentation, and offers “virtue from error” as a personal operating principle.
- •Innovation correlates with comfort failing publicly and iterating quickly
- •Highly curated paths can reduce playfulness and experimentation skill
- •“Virtue from error”: turn mistakes into progress rather than shame
- •Extraordinary career impact often requires extraordinary, non-linear bets
- 58:25 – 1:02:35
Joy, intensity, and doing your best: finding the work you can’t believe is paid
Sam and Lenny discuss career pragmatism and privilege, while still advocating attention to energizing moments and strengths. Sam suggests watching for what others praise you for, leaning into it with intensity, and doing the best job possible even in imperfect roles.
- •Stop doing miserable work as soon as you realistically can
- •Look for signals: what work feels surprisingly natural and high-impact?
- •Bring your unique strengths to any role; intensity can be an advantage
- •Doing your best where you are builds momentum and opportunity
- 1:02:35 – 1:09:28
AI as the next platform shift: apps become features of AI, and pixels get “free”
Sam lays out his AI worldview: models matter, but the bigger disruption is building systems around them—state, orchestration, tools, agents, and dynamic UI. He compares the moment to early web apps and predicts a major re-architecture of software over the next decade.
- •“AI isn’t a feature; your product is a feature of AI”
- •Transformative value comes from AI-native products that can’t exist otherwise
- •AI may make “pixels free,” similar to how the internet made distribution cheap
- •Future: intention-driven agents, personalization, multimodal dynamic interfaces
- 1:09:28 – 1:13:09
How to approach AI without falling behind: choose a goal, then grind toward it
Sam rejects vague advice to “just play with it” and recommends goal-driven learning. Pick a meaningful task in your domain, try to automate or augment it with AI, and learn by pushing through obstacles—while accepting the era is inherently exhausting to track.
- •Set an explicit learning target (a real job task), not unstructured tinkering
- •Be stubborn about reaching the goal; the struggle reveals what matters
- •The flood of use cases is itself a signal of a true platform shift
- •Zero-to-one insight is rare; one-to-many optimization will accelerate rapidly
- 1:13:09 – 1:17:33
Why Microsoft feels innovative right now: leadership energy, humility, talent, and bets
Sam attributes Microsoft’s momentum to Satya Nadella’s leadership style—empathy, energy, and engagement—plus a humble execution culture and deep technical bench. He also notes the role of strategic risk-taking and luck, including the OpenAI partnership bet.
- •Satya’s leadership: raises organizational energy without destructiveness
- •Culture emphasizes humility and unglamorous hard work
- •Extraordinary talent density and deep experience across teams
- •Big strategic bets (and luck) matter alongside culture and execution
- 1:17:33 – 1:27:50
Lightning round and close: books, shows, interview questions, and blood sausage
In the lightning round, Sam shares influential books, favorite shows, and a provocative interview question designed to test how candidates react to “impossible” problems. The episode ends with a quirky story: selling both a company and 200 pounds of blood sausage to Google, plus where to follow Sam’s writing.
- •Book recs: Invisible Cities, The Wasp Factory, Where Good Ideas Come From
- •Favorite shows: Slow Horses and monster/sci-fi comfort viewing
- •Interview test: resilience and reasoning under seemingly impossible prompts
- •Find Sam via Sunday Letters from Sam (Substack) and LinkedIn
