🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

Cooperation on AI safety is necessary but also comes with potential risks. In our new paper, we identify technical AI safety areas that present comparatively lower security concerns, making them more suitable for international cooperation—even between geopolitical rivals. 🧵

🚨 Excited to share my new paper with Atoosa Kasirzadeh “Characterizing AI Agents for Alignment and Governance” We identify four key dimensions of AI agency—autonomy, efficacy, goal complexity & generality—and use this analysis to construct “agentic profiles” for a range of AI agents.

This looks like a really helpful contribution!

Iason Gabriel 1/ We identify four key dimensions of AI agency: autonomy, efficacy, goal complexity, and generality. For each dimension, we propose a set of graded categories. We then use our analysis to construct "agentic profiles" for four different types of AI systems

Do you want to think sharply about AGI? Well, first you need to know what AI agents are. For that, Iason Gabriel and I have an answer for you, with tables and visuals: arxiv.org/pdf/2504.21848

Cool paper alert

Some Japanese coverage 😁

Check out this amazing tool that Andrew Maynard made! It allows you to plot and visualise the dimensions of AI agency contained in our new paper "Characterizing AI Agents for Alignment and Governance" with Atoosa Kasirzadeh: fvture.net/agentPlot/