Stories by erwald

Where will China get its compute in 2026?

Why securing AI model weights isn't enough

The case for paying whistleblowers to report on export violations

A Whistleblower Incentive Program to Enforce U.S. Export Controls

Not a Meat Eater FAQ

I'm Switching into AI Safety

Midihum: An ML-Based MIDI Humanizing Tool

Attention on Existential Risk from AI Likely Doesn't Distract from Current Harms

10 Years of Earning to Give

A List of (Almost) Every Single Chinese Tofu

A Nordic revolt against 'ugly' modern architecture

Tentative practical tips for using chatbots in research

Alpha

Misaligned AI constitutes a growing public health threat

Yes, avoiding extinction from AI is an urgent priority

The Prospect of an AI Winter

Power Smoothie

Against LLM Reductionism

How to Get Hold of a Research Paper

Measuring Good Better

1
2