47

The Garden

A collection of essays, notes, short films, and half-baked explorations I'm always tending to.

Topics

Why we should vaccinate wild animals - Works in Progress Magazine

Can You Poach A Frontier Lab?

Why Europe doesn’t have a Tesla - Works in Progress Magazine

Dario Amodei — Policy on the AI Exponential

Europe 2031 — What getting AI wrong means for us

What does your chatbot use actually cost?

How Long Until AI Doesn’t Need Humans?

Meanders

the html review 05

The world’s most complex machine - Works in Progress Magazine

When AI builds itself

Don’t sell yourself short

Teaching Claude why

Natural Language Autoencoders

Google DeepMind Workers Vote to Unionize Over Military AI Deals

LawZero | The Scientist AI: Safe by Design, by Not Desiring

A call for more specific and numerate climate communication

Claude Mythos and misguided open-weight fearmongering

Claude Mythos: The System Card

Anthropic's Pause is the Most Expensive Alarm in Corporate History [Fiction] — LessWrong

Integrating Agricultural Development and Methane Mitigation as part of the EU Livestock Strategy

Europe’s SMR moment has arrived — now comes the hard part

On restraining AI development for the sake of safety - Joe Carlsmith

Claude Opus 4.6: System Card Part 2: Frontier Alignment

Anthropic Drops Flagship Safety Pledge

Letter from Utopia

ML: Time to Embrace Version Control | Luca Palmieri

Dario Amodei — The Adolescence of Technology

Building Brains on a Computer

Claude's Constitution

What sort of post-superintelligence society should we aim for?

Is China Serious About AI Safety? | AI Frontiers

AI Will Be Your Personal Political Proxy | AI Frontiers

Thoughts on anti-AI-safety arguments

Thoughts on anti-AI-safety arguments

My personal reflections and thinking for why AI safety is important, and my responses to what i think are the best counter-arguments.

December 2025

- Probability and Statiscs course. - Programming and Problem solving course, we have...

How do we solve the alignment problem? - Joe Carlsmith

The AI Was Fed Sloppy Code. It Turned Into Something Evil. | Quanta Magazine

Towards Guaranteed Safe AI: A Framework for Ensuring Robust and Reliable AI Systems

Will the Humanities Survive Artificial Intelligence? | The New Yorker

Avoiding an AI Arms Race with Assurance Technologies

A Brief History Ethos of the Digital Garden

Post 34: Learning how to learn

Here’s How To Share AI’s Future Wealth

in defense of pretension.

Montuzzo

A film montage focusing on cinematography and music, set in the abandoned ruined city of Montuzzo in northern Italy

Montuzzo Suite

Original composition inspired by Gustav Holst's 'Venus', accompanying the short film of the abandoned ruined city

Artificial Intelligence: Impact on the Individual and Legislation

Artificial intelligence affects us all, from the algorithm in Facebook to AI systems that help discover new medicines. Recently, we've all heard about ChatGPT, and many are frightened by this rapid development. But what are the actual problems, and how do we solve them?