My work has been discussed in the Guardian, the Times, and in my own piece for the Economist. I also co-organised the workshop on Post-AGI Civilizational Equilibria.

Some research highlights:

Gradual Disempowerment (tweet thread) — a paper on how the proliferation of advanced AI might disrupt human influence over key societal systems by naturally and predictably displacing humans.

Decomposing Agency — a post on how AI models might contribute different features to agent-like systems rather than simply being agents themselves.

Limitations of Agents Simulated by Predictive Models — a workshop paper on theoretical challenges to eliciting agents from models that were pretrained offline on prediction.

“AI for Societal Uplift” as a Path to Victory - a post where I argue that we should think about using AI advances to radically improve humanity’s collective reasoning ability, so that we can navigate the risks it will unlock.