🔥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

It's been great fun to work on this project and figure out what good abstractions look like here - I hope people can find it useful for eval development!

You can find Inspect here: ukgovernmentbeis.github.io/inspect_ai/

Our 2024 Winter Fellowship recently ended. Here we share summaries of the projects our Winter Fellows worked on during their three months with us: governance.ai/post/winter-fe…

My main concern as global AI governance conversations go on is, how do we ensure negotiations are in good faith? Bargaining powers and leverage is very crucial in regime development. Reflecting on this incredible piece by Judd Devermont

Most real-world AI applications involve human-model interaction, yet most current safety evaluations do not. In a new paper with Saffron Huang Lama Ahmad لمى احمد Markus Anderljung, we argue that we need evaluations which assess human-model interactions for more accurate safety assessments 🧵

Amazing Sofya Lebedeva 🙌🏽🙌🏽

Just a reminder this advice of ‘you don’t need to go to school’ is not for people like US! Go to school!!!!

Amazing Dr Kerry McInerney (Mackereth) 🙌🏽🙌🏽

LawAI is hiring! bit.ly/law-ai-open-po… The Institute for Law & AI (LawAI) is growing substantially! We’re hiring at all levels of seniority, across our teams, including roles in operations, programs, research, and consulting. If you’re looking for a supportive,

New research memo! The AI Safety Institutes (AISIs) are poised to assume an increasingly significant role in the governance of advanced AI. We recently held an expert workshop to explore the roles AISIs can play. Find out more here Oxford Martin School oxfordmartin.ox.ac.uk/publications/a…

Apply for the GovAI 2025 Winter Fellowship! Fellows join GovAI to conduct research on a topic of their choice, with mentorship from leading experts in the field of AI governance. The deadline to apply is Sunday 11 August 2024. governance.ai/post/winter-fe…

We're excited to launch Inkuba LM 4.0B, the first small language model for IsiZulu, Yoruba, Hausa, Swahili, and IsiXhosa. InkubaLM is specially designed to support African languages and promote linguistic diversity in AI. Try it out: lnkd.in/dA6aSPCf #InkubaLM

New essay out this week on the Collective Intelligence Project blog, co-written with Saffron Huang: "Predistribution over Redistribution: Beyond the Windfall Clause" Check it out! cip.org/blog/predistri…

Belated sharing of a T20 Policy Brief I contributed to lead by the wonderful Jai Vipra on 🌟 "Governing Computational Infrastructure for Strong and Just AI Economies" 🌟 part of the T20 Task Force on Inclusive Digital Transformation. Read more rb.gy/5tbjyc

Here you can listen to Dasha Pruss and I talk about ideas and findings from our #FAccT2024 paper “Evidence of What, for Whom?”, and the socially contested role of algorithmic bias with great host Alix Dunn ! Thanks for the opportunity

Check out our new research on the North-South divide of AI compute with Vili Lehdonvirta and Zoe Hawkins covered by one of my favourite journalists, Billy Perrigo, in TIME 🎉 time.com/7015330/ai-chi…

How can nations achieve a shared, scientifically rigorous understanding of AI risks? Check out our new report that examines the challenges and opportunities! oxfordmartin.ox.ac.uk/publications/t…

Our director Robert Trager is interviewed by Gavin Esler in the latest episode of This Is Not A Drill - with Gavin Esler, exploring the risks and opportunities of advanced AI systems. Oxford Martin School

📢 Join OCPL as a Programme Associate! 📍Oxford, UK | Part-time (Oct 2024-Jul 2025) 💡Role: Support implementation of OCPL projects, including project management for the Lab's flagship fellowship programme. Full posting: oxfordchinapolicylab.com/program-associ…