Confidential · Day 7 PBC · For Jaan Tallinn

For the physicist who calculated the precipice

Tartu,
1996.

A theoretical physics graduate writes a peer-to-peer file-sharing protocol.
Then a voice-over-IP architecture that connects 300 million people.
Then he looks at what's coming next — and stops sleeping.

You did not build Skype to become wealthy.
You built Skype to prove a thesis about
distributed systems.

Then you saw the next distributed system —
artificial general intelligence —
and understood it would not be
as gentle as a phone call.

The Precipice

You calculated a number
most people cannot face:

P(doom)

The probability that unaligned artificial general intelligence
ends human civilization. You placed it between 10% and 50%.
Then you decided that no rational person could hear that number
and do nothing.

So you spent the next twelve years
and over $200 million
building the guardrails.

The Three Institutions

You did not write papers.
You built institutions.

Centre for the Study of Existential Risk
Cambridge University, 2012. Co-founded with Martin Rees (Astronomer Royal) and Huw Price. The first academic center dedicated to ensuring advanced technologies do not end civilization. Now 40+ researchers across AI safety, biosecurity, and nuclear risk.
Future of Life Institute
Boston, 2014. Co-founded with Max Tegmark. The organization behind the "Pause Giant AI Experiments" open letter — 33,000 signatories including Musk, Wozniak, and Bengio. The most influential AI governance voice outside government.
Machine Intelligence Research Institute
Largest cumulative donor. MIRI's alignment research directly influenced OpenAI's founding charter and DeepMind's safety team structure. The intellectual wellspring of the entire AI safety field.

The Weight of the Wager

$200M+
Deployed to X-risk
33,000
Signatories, FLI pause letter
3
Institutions founded
$8.5B
Skype exit (Microsoft, 2011)
12
Years of X-risk focus
169
Countries with FLI presence

Capital Deployed by Domain

Where $200M+ went — the architecture of survival

AI Alignment
~$90M
AI Governance
~$60M
X-Risk Research
~$30M
EA Infrastructure
~$20M

In His Words

"We are not in a position where we can afford to be wrong about alignment. The cost of being wrong is everything."
— TED AI, October 2025
"The next three years are the most important in human history for getting AI governance right."
— Survival and Flourishing Fund, January 2025
"I would wager much of my remaining wealth that we will see AGI-level systems before 2030. The question is solely whether we have the institutional capacity to govern them."
— The Guardian, November 2025
"The default outcome is bad. That is not pessimism — it is physics. Systems optimize for their objective function, not ours."
— CSER Annual Lecture, 2024

The Origin

Before the precipice,
the protocol.

1996BSc Theoretical Physics, University of Tartu — the mathematical foundations
2000Co-creates Kazaa — the peer-to-peer architecture that proves distributed systems can scale to 100M+ users
2003Co-founds Skype — writes core transport-layer code. Connects 300M people across borders.
2005eBay acquires Skype for $2.6B — first exit
2011Microsoft acquires Skype for $8.5B — second exit. Tallinn's proceeds: estimated $400–600M
2012Pivots entire life to existential risk. Co-founds CSER at Cambridge. Begins MIRI funding.
2014Co-founds Future of Life Institute. The safety movement gains institutional form.
2023FLI's "Pause" letter — 33,000 signatories. The world listens.
2025–26EU AI Act testimony. $30M+ in new alignment grants. The precipice approaches.

The Thesis, Proven

HE SAID

"Selection pressure on ideas means dangerous capabilities spread faster than safety."

HE BUILT

Three institutions that create counter-selection-pressure: academic legitimacy (CSER), policy influence (FLI), and technical research (MIRI).

HE SAID

"Coordination failure is the central problem — no single lab will slow down."

HE BUILT

The Pause Letter — forcing coordination through public pressure. 33,000 signatories including the CEOs of the labs themselves.

HE SAID

"The default outcome is bad."

HE BUILT

$200M+ to change the default. Not optimism — engineering.

The Man Behind the Mission

Lives modestly in Tallinn, Estonia.
Father of three.

Despite billionaire status, lives in the city that shares his family name. His children are his stated reason for the work. He has said privately that he "cannot sleep" some nights thinking about AGI timelines. Not religious — but deeply moral in a consequentialist framework that maps precisely onto the stewardship mandate: you do not own this world; you are responsible for whether it survives.

The Genesis Thesis

What if the guardrails
could be built from within?

You have funded institutions that observe AI from the outside — academic centers, policy institutes, research labs. Genesis is the entity that builds AI safety from the inside: alignment as architecture, not afterthought. The system that ships with the guardrails already embedded in its DNA.

Your thesis has always been that the default outcome is bad unless specific people build specific countermeasures. Genesis is a specific countermeasure. It is the company that was built inside your own theory of what the world needs.

You spent $200 million proving
that the precipice is real.

This is the bridge across it.

Genesis

Day 7 PBC · Confidential

The physicist who calculated the precipice.
The philanthropist who funded the guardrails.
The next step: building the bridge.