Strange Cities

The crucible

How I think about the situation with AI

May 2 • 

Owen Cotton-Barratt

7

1

April 2025

Disempowerment spirals

Notes on a likely mechanism for existential catastrophe

Apr 10 • 

Owen Cotton-Barratt

 and 

Raymond Douglas

1

March 2025

Knowledge, Reasoning, and Superintelligence

What makes people good at solving novel problems?

Mar 26 • 

Owen Cotton-Barratt

1

AI Tools for Existential Security

(From a paper coauthored with Lizka Vaintrob.)

Mar 17 • 

Owen Cotton-Barratt

1

November 2024

The Choice Transition

On the emergence of history’s reins

Nov 18, 2024 • 

Owen Cotton-Barratt

 and 

Raymond Douglas

1

A brief history of the automated corporation

Looking back from 2041

Nov 4, 2024 • 

Owen Cotton-Barratt

October 2024

AI safety tax dynamics

Earlier AI capabilities — research and coordination — could help us to navigate later safety problems. And the highest risk probably comes in the era…

Oct 23, 2024 • 

Owen Cotton-Barratt

Safety tax functions

What's the relationship between "looking for a solution" to a safety problem, and "paying a safety tax"? By considering tax requirements as a function…

Oct 19, 2024 • 

Owen Cotton-Barratt

September 2024

AI, centralization, and the One Ring

Articulating the natural concerns with amassing power in a single AGI project

Sep 11, 2024 • 

Owen Cotton-Barratt

2

July 2024

The last era of human mistakes

Author’s remark, four months later: in retrospect I am vaguely dissatisfied with this piece.

Jul 24, 2024 • 

Owen Cotton-Barratt

1

Decomposing Agency

People sometimes say that AGI will be like a second species; sometimes like electricity. Both may have elements of truth. We need concepts which let us…

Jul 11, 2024 • 

Owen Cotton-Barratt

 and 

Raymond Douglas

7

June 2024

AI takeoff and nuclear war

Drivers of all-out war, and strategies to mitigate the risk

Jun 11, 2024 • 

Owen Cotton-Barratt

6