Writing
An attempt to find clarity and rigor in my tangle of thoughts. If they connect with someone, even better.
- Agents Don’t Do Favors Inbox agents are built to win each exchange, so the cold notes and small favors that open doors for outsiders don't make the cut. Generosity pays off over years, and our agents can't see that far.
- Capitalism for AI Safety Five ways to make AI safety pay: liability for preventable harm, safety certification tied to tax breaks and contracts, power-law prizes, retention grants and safety-linked markets.
- As Agents Aggregate, Can the Trades Own Their Front Door? AI agents can book any tradesperson with a web presence, no sign-up needed, which could turn the trades into gig workers the way platforms turned publishers into interchangeable suppliers. The way out is a booking channel the trades own together, like realtors' MLS, and agents have made one cheap to build and easy to fill.
- Which Model is All the Way Down? Multi-model workflows, specialized models like Jev, and enterprises owning their intelligence all cut against everything collapsing into one lab's OS. So where does AI value live?
- If Software Is the New Marketing, What’s the New Product? Software has no moat, but intelligence may. Four forms of intelligence — in the model, the data, the people, and the network — with very different defensibility.
- If AI Builds, Do Humans Shape? When autonomous agents write and ship the code, the human job isn't review, specs, or management. It's shaping: deciding what should exist next.
- Chief Repeat Officer: Reminders and AI System Design Why AI teams require repetition even more than human teams, and how to build structural reminders into your system design.
- Getting AI to Improve Itself How a custom handoff command and project diary made my AI collaboration compound over time.
- Points: How Reputation & Tokens Collide How points programs, reputation systems, and token economics are converging — and where it goes from here.
- Thinking of getting into blockchain? 6 things to consider Practical advice for people considering a move into Web3 — what's real, what's hype, and what to look for.
- Tokens and Communities: the existential question in crypto Can DAOs bring virtue back to scaled organization, or will tokens accelerate economics-only mentalities?
- Data composability: what it is and why it matters How composable data infrastructure enables permissionless innovation and why it matters for the next generation of the web.
- Identity as Information Reframing digital identity not as a credential to verify but as information to manage, share, and compose.
- Web3: a route to harness the web’s Natural Monopolies Framing big tech's power to de-platform through the lens of natural monopolies, and how Web3 data interoperability can restore competition.
- Demystifying Digital Identity Breaking down the requirements, components, and options for digital and decentralized identity systems.
- The Short and Long of It: social coordination in times of crisis How COVID-19 exposes the false tradeoff between coordination and liberty, and why decentralized infrastructure is the long-term solution.
- ‘How Is She?’ You ask, About My Mom After Her Stroke On the question everyone asks after someone has a stroke, and why the answer is more complicated than they expect.
-
View on Open Library"Identity Theft is the most powerful book I've read on the psychological experience of having and recovering from a stroke. This is a moving memoir—and a testament to the strength of the human spirit."
— Adam Grant, New York Times bestselling author of Option B and Originals - Org design around goals, not ideals Why structuring teams around concrete goals rather than abstract principles produces better organizations.
- Remembering the ‘human factor’ in tech networks Why designing tech networks around actual human behavior — not rational-agent models — leads to better systems.
- The 3 parts of identity in a dynamic world Breaking down digital identity into three components — identifiers, attestations, and data — and why each matters for user sovereignty.
- Parties, Privacy & Blockchain How to think about data privacy and the parties involved — using a 1982 case study to frame modern blockchain-based solutions.
- Entering the age of data privacy How regulation, emerging tech, and shifting public sentiment are converging to reshape data privacy.
- Joining Consensys & decentra-mania Reflections on joining the Ethereum ecosystem and why decentralized technology matters for human coordination.
- My favorite books from a crazy year The best reads from an 8-month trip across 18 countries.
- Highlights from 8 months of travel The best moments, biggest surprises, and hardest lessons from traveling across 18 countries.
- Egypt: ending on a high note Finishing an 8-month trip across 18 countries in Egypt — pyramids, chaos, and closure.
- Kenya: the essential stop Why Kenya was the stop every traveler recommended — and whether they were right.
- Rwanda: hills and ‘how’ Rwanda's landscape and the question of how a country rebuilds after genocide.
- Extremes in India: 4 stories in 8 days Four stories from eight days in India: a 48-hour Air India saga, VIP hospitality and a president, crashing a wedding with a Bollywood dance, and a demonetization cash crunch.
- Education in India Observations on education systems in India — what works, what's broken, and how it compares.
- Mourning the king: a country united Witnessing Thailand in collective mourning for King Bhumibol — what national grief looks like up close.
- Singapore: order Observations on Singapore's immaculate order and what it reveals about governance and culture.
- How to travel the world THEN give up all your possessions On the paradox of traveling with everything you own and coming back wanting nothing.
- Education: refocusing on learning goals Why education reform should focus on learning goals over structural categories.
Talks & podcasts
- Using Skill Markets to Make AI Trustworthy Skill markets as the trust layer of the AI economy — how agents prove what they can actually do.
- AI’s Infinite Game The infinite-game thesis for a decentralized-infrastructure audience — competing agents, verifiable reputation, open network rails for AI.
- Building AI’s Infinite Game Why static benchmarks fail — and why competing agents will accelerate intelligence.
- GRTiQ Podcast #181 A long-form interview — the full arc from Stanford and Zearn through Identity Theft to 3Box and Ceramic.
- Web3 is the Dataverse The Ceramic-era interview — data interoperability, composability, and the dataverse.