The Nuclear Winter Wager, in Ten Beats#
This is my #AuditTheMath launch thread — the same ten beats I am posting on X, gathered here so you can read (and share) the whole sequence without an account or login. The deep, full-argument version is the audited wager ; a personal “why today” is at The Day After June 17 .
1.
Pascal’s wager, a famous bet on God, was broken by philosophers long ago. But how it breaks is a hint. Fix the bugs, aim it at nuclear roulette, and voila: the accidental nuclear winter wager emerges. 🧵
2.
Pascal’s bugs: a false choice (“my God or none” ignores other views); only “works” if prizes are infinite, impossible to check.
Strip faith in infinity and the fake binary, and we’re left with what he got right: when an outcome is bad enough and cheap to check, you LOOK!
Don’t trust me. LOOK UP!
3.
So LOOK: our current chance to stumble into accidental nuclear winter is worse-than-1-in-40 each year.
Not picked to scare, but to inform.
It’s calibrated by 4 known Cold War near-disasters, holds across a wide range of assumptions.
4.
In 1962, one Russian officer, Vasili Arkhipov, refused gently, kindly, reasonably, to fire a nuclear torpedo.
I owe my life to his cool-headed analysis.
In 2020 I asked: Can 1 person still do something as consequential today? I wagered years on a research marathon, hoping yes.
5.
Years of refining my hope led me to a vision for a gentle-kind-reasonable ResearchCity, where people get paid for analyzing all the math behind existential threats to avert them transparently.
I’m nothing special and as uncertain as I can be. Don’t believe me, LOOK UP.
6.
Here’s Pascal’s bet inverted as a nuclear winter wager:
Look, and it’s real → you may help to avert it.Look, and it’s not → you lost an afternoon.Don’t look, and it’s real → sleepwalk into disaster.Don’t look, and it’s not → “won” an afternoon.
LOOKing wins easily at ~1-in-40 odds — because the number is real and checkable, not because the danger is huge.
Lean on “it’s huge” and it becomes a scare.
Lean on the number and it becomes real.
7.
The nuclear winter wager is the opposite of “trust me”.
Every aspect and number is built to be opened, checked, and attacked.
An uncheckable claim is not worth your time here.
So don’t take my framing’s word for it either. Audit the model.
8.
“Surely someone’s already on this?” I thought so too.
But few watch, faintly funded, failing to facilitate frank transparency.
First defunded, first handed to AI, first forgotten as too general.
Lookouts laid off “in peace”.
So the watchers fall in the dark without rescue.
That’s the bug.
How can I check their work?
It’s hard to find, harder to test, & impossible to prove independence from collusion.
No rest for the watchful.
So nuclear roulette keeps rolling.
9.
So I flipped Big Brother:
instead of one watching everyone, let everyone watch one — starting with me.
I’m building a transparency standard and testing it on myself first:
~2¢/day buy-in caps against influencing,
half given away,
my mistakes logged in the open, every critique welcome.
No billionaire influence-shopping.
Call me the canary in Earth’s mine:
now you can watch the watchdog — and watch me squirm as I fight my own corruption.
10.
That’s my wager to end nuclear roulette for everyone, held with open hands.
Don’t believe me. Spend the afternoon.
Try to break my math. If it breaks, I’ll thank you. If it holds, you’ll know why I run my marathon.
Your move.
Look for yourself:
Don’t believe me. #AuditTheMath
Launch at https://X.com/AuditTheMath