Public Goods, Private Rails.

Seeking common ground

Progress takes more than a breakthrough.

A shared inquiry byDorothy Chou, Sandro Gianella & Florian Maganzaon technology, public purpose and the work between them.

Our studies and careers have intertwined for more than fifteen years. Lately, our conversations have become more frequent, and more impatient. After two decades in technology, we keep coming back to one question:

What needs to be built around a breakthrough for it to benefit society broadly?

Our starting hypothesis is that progress depends on the work around the technology as much as on the technology itself.

AI can help identify a promising drug candidate. Turning it into a treatment people can access still takes laboratory research, clinical trials, regulatory approval, manufacturing and a health system able to deliver it. The discovery is only a beginning.

Companies, philanthropy and public institutions each have a role. Too often, no one is responsible for joining the pieces up.

That is the gap we have spent our careers navigating. We are putting our thinking into the open to explore how to bridge it. We would love you to read and think along with us.

We think of this as a shared garden. Commercial investment, philanthropy and public institutions each tend a part; our interest is in the garden as a whole, and how it can flourish for the wider community.

Read our introduction
Open garden paths and a bridge connect a shared landscape.
The common gardenOriginal illustration for the series

Published · 4 min read

Start here

Our shared starting point: what surrounds a breakthrough, and why we want to think it through together.

Read the introduction

Get an email when the next essay is published

  1. 01
    A cart carries young plants from a walled nursery toward shared garden plots: cultivation made useful beyond its place of origin.
    Beyond the garden wallOriginal illustration for the series
    Published now18 min read · including foreword

    Public benefit

    Turning Private Technology Breakthroughs into Public Goods

    How can a breakthrough inside a private company become something many more people can use?

    Read essay
  2. 02
    A stone footbridge connects a leafy garden path to a quiet village courtyard.
    The crossingOriginal illustration for the series
    Not yet published

    Public services

    Why Cutting-Edge Technology Isn’t Making It into Public Services

    What stands between a promising technology and an NHS service that can put it to work?

  3. 03
    A gravel path forks around an olive tree, passing through two open garden gates.
    Room to moveOriginal illustration for the series
    Not yet published

    Sovereignty

    Agency, Not Ownership

    How can countries gain the power to use and shape AI without owning every part of it?

  4. 04
    Prepared beds, trellises and a glasshouse make it possible to invest in the coming harvest.
    Preparing the groundOriginal illustration for the series
    Not yet published

    Demand

    Buying the Future into Existence

    What becomes possible when governments commit to buying an outcome before the market exists?

  5. 05
    Young fruit trees supported by timber stakes grow beside mature trees and a stone irrigation channel.
    The patient orchardOriginal illustration for the series
    Not yet published

    Institutions

    Technology the Market Won’t Deliver on Its Own

    Who builds the shared tools, evidence and institutions that useful technology needs?

A gardener’s workbench opens onto a shared garden. A young tree, well-used tools and an open seed packet sit beside three small pots, ready for something new to grow.
The shared workshop

Choose a pot to plant a seed.

A shared tool

The work between us.

The work between institutions is also work between people. We like Joseph Beuys’s two-handled spade as a companion to these essays: a shared tool that asks its users to find a shared rhythm.

Joseph Beuys, Wuppertal, 1965.
Photograph © Ute Klophaus.

Behind the essays

Why we keep
coming back to this.

From our student days at the University of Toronto, Sussex and Georgetown, our paths have crossed through two decades in technology. We have built teams, programmes and partnerships through periods of extraordinary growth, working at the intersection of public policy, science and business.

Our work has taught us to look across disciplines: at how a scientific advance, a business model, a public institution and the people affected by a decision fit together. Our friendship gives us room to compare what we have learned, challenge one another and follow a question further than any of us would alone.

As operators

Inside growing companies.

At DeepMind, Dorothy built the public engagement team around advances including Nobel-winning AlphaFold. At Stripe and OpenAI, Sandro built policy and partnership functions as the companies scaled into new markets. Florian built Google.org’s European presence and impact measurement work, then Frontier’s communications at Stripe.

These experiences taught us how much depends on the work around a breakthrough: earning trust, establishing evidence, shaping policy and connecting people who need one another.

Alongside founders

Learning by backing builders.

As angel investors, venture partners and scouts, we also learn from founders working on problems that matter to us. Between us, our investments include:

Founders bring us questions about evidence, regulation, partnerships and buyers. Those conversations keep this inquiry grounded in the work of making things useful.

Dorothy Chou

Dorothy Chou

Dorothy builds connections between scientific discovery and public life. At DeepMind, she built teams responsible for communications, public engagement and government relations as the company grew, including around AlphaFold, the AI system recognised with the Nobel Prize in Chemistry, and its early release to scientists working on neglected diseases.

She designed and launched Google.org’s AI for Science fund and now chairs UCLPartners, where NHS organisations, universities and industry work together to improve care. Earlier roles at Google, Uber and Dropbox included creating Google’s Transparency Report and leading a criminal justice reform campaign at Uber.

She has backed founders as an angel investor and as a Venture Partner with Norrsken VC and Ada Ventures, and also scouted for Atomico and Giant Ventures. She studied at Georgetown and Oxford. A question running through her work is how a discovery becomes something people can use—and have a voice in shaping.

Sandro Gianella

Sandro Gianella

Sandro is interested in the conditions that allow a new technology to spread. His work at Google, Stripe and OpenAI has shown him how much depends on the policies, partnerships and institutions around it.

As Stripe’s first public policy hire globally, he built its EMEA function and helped open new markets. He joined OpenAI’s founding public policy team, helped negotiate AI safety commitments at Bletchley Park, and now works on growth across seven priority markets.

He backs founders as an angel investor and Venture Partner at Sunlit Ventures, and advises through Atomico’s Expert Network. He studied at the University of Toronto and the Hertie School of Governance in Berlin, and is completing the ETH Zurich–St. Gallen Executive MBA. He keeps asking how the choices of companies, governments and investors shape what a technology ultimately makes possible.

Florian Maganza

Florian Maganza

Florian focuses on new ways to fund and scale work that delivers public benefit. As Google.org’s first hire outside the US, he established its presence and strategy in EMEA and led Google’s efforts to connect refugees with vital information.

After moving to the US, he led a grant portfolio supporting the use of technology to match low-income job seekers with work and help them develop new skills. He then built Google.org’s impact measurement and learning function, using insights from hundreds of nonprofit organisations to make better funding decisions. At Stripe, he built Frontier’s communications function, helping explain which carbon removal approaches its buyers backed and why.

He studied at Sussex University, the London School of Economics, and the College of Europe. A question he returns to is how you build a market for a product that’s needed but doesn’t have a natural customer.

We write here in our personal capacities. The views expressed are our own and do not represent those of any company or organisation with which we are affiliated. This publication is an independent collaboration and is not produced on behalf of, or endorsed by, those organisations.

Stay in the conversation

Keep thinking with us.

Occasional essays, reading notes and questions from Dorothy, Sandro and Florian.

Only for these updates. Your address stays private. Unsubscribe whenever you like. How we look after your information.

Something we should read? A question or a different view? Write to us

A curious footnote

A spade for two.

Joseph Beuys · Gemeinschaftsspaten
Community Spade, 1964

Two spades in a museum display case. Each heart-shaped metal blade has two long wooden handles that spread apart, ending in T-shaped grips.
Two spades, each with two handles. View larger ↗ © Hessisches Landesmuseum Darmstadt. Photograph: Dennis Haustein, 2025; VG Bild-Kunst, Bonn 2025. Museum collection ↗

Beuys made a spade with two handles, then used it in a performance at Galerie Parnass in Wuppertal in 1965. Ute Klophaus photographed him holding its heart-shaped blade at his chest, with the handles reaching upwards.

“Two handles on one spade signify a special kind of compound action for people working the earth together.”

— Joseph Beuys, quoted by Dia Art Foundation

What draws us to it is the little practical puzzle: a shared tool asks people to find a shared rhythm. Having the tool is only the beginning; something has to happen between the people holding it. We like that as a companion to these essays, and as a reminder to leave room for a little humour.

Opening-page photograph © Ute Klophaus. Image excerpt from MoMA’s Thinking Is Form catalogue (1993), p. 52, fig. 31 ↗.