1 May 2026Inkhaven
Linkpost: Sanity-Checking 'Incompressible Knowledge Probes'Writing
Essays, research notes, and stories — including the Inkhaven 30-day run.
29 April 2026Inkhaven
Q&A28 April 2026Inkhaven
Attachment Theory Is Extremely Cool and Useful27 April 2026Inkhaven
From Output to Connection26 April 2026Inkhaven
Why Are Anglophone Countries Unhappier Than Their European Counterparts?25 April 2026Inkhaven
The Great Smoothing Out24 April 2026Inkhaven
Do Emergently Misaligned Models Believe What They Say?23 April 2026Inkhaven
Whatever I Was Doing in 2020, I Should Do Again (Continued)22 April 2026Inkhaven
Whatever I Was Doing in 2020, I Should Do Again21 April 2026Inkhaven
Giving AI a Childhood20 April 2026Inkhaven
Contributing to Technical Research in the AI Safety End Game19 April 2026Inkhaven
People That Inspire Me18 April 2026Inkhaven
Which Information Sources I Listen to the Most17 April 2026Inkhaven
Do Models Know They're Lying When Claiming Fake Identities?16 April 2026Inkhaven
You Can Just Build Things with Claude Code15 April 2026Inkhaven
A Case for Persona Robustness as a Research Area14 April 2026Inkhaven
Characterising the Views on Safety from Frontier AI Labs: OpenAI13 April 2026Inkhaven
Characterising the Views on Safety from Frontier AI Labs: Anthropic and DeepMind12 April 2026Inkhaven
Is the Adult in the Room with Us Right Now?11 April 2026Inkhaven
How to Make a Problem Disappear10 April 2026Inkhaven
How South Africa's Electricity Catastrophe Was (Mostly) Fixed9 April 2026Inkhaven
This Is Not Financial Advice7 April 2026Inkhaven
Revisiting GSM-Symbolic: Do 2026 Frontier Models Still Fail at Confounded Grade School Math?7 April 2026Inkhaven
Spend Money to Buy Opportunity6 April 2026Inkhaven
How Does an Agent with Multiple Goals Choose a Target?How a maze-solving RL agent internally represents and switches between multiple sequential goals, discovering spatial gating through negative activations.
5 April 2026Inkhaven
Spectator SportWatching a cybersecurity competition where the most promising minds have become spectators to their own tools.
4 April 2026Inkhaven
The Garden (Complete)Two angels debate whether paradise requires suffering. One builds elegant systems; the other asks why they can't do better.
3 April 2026Inkhaven
The Garden (Part 1)Two angels tasked with building paradise discover that creation demands uncomfortable trade-offs.
2 April 2026Inkhaven
The Physics of Great StorytellingHow the best fiction earns its weight by letting the world do the work — using Hunter X Hunter and A Wise Man's Fear as examples.
1 April 2026Inkhaven
How did the United States hand over its future to corporations?Tracing how a single memo by a corporate lawyer led to the systematic dismantling of the policies that built America's middle class.
December 2025
Whole Brain Emulation as an Anchor for AI WelfareNovember 2023
Why pursue conceptions of agency for AI safety2023
Vipassana Meditation and Active Inference: A Framework for Understanding Suffering and its CessationJanuary 2022
Lessons from my first 10 day Vipassana