Dear Apart Community,

Welcome to the first edition of our new newsletter - Apart News!

At Apart Research there is so much brilliant research, so many events, community updates, and more to share that we have decided to start this new newsletter to do just that.

Our community of researchers is so important to us and Apart News will reflect that. This newsletter will usually include this type of content:

• Research coming out of Apart Lab;
• Hackathon and Sprints results;
• AI Safety research we have found interesting;
• Excerpts from our newest blogs;
• Highlights from our upcoming 'Researcher Spotlight Series' interviews.

For those of you who need a refresh: Apart is a global collective conducting impactful AI Safety research - a decentralized laboratory mobilizing researchers.

AI Safety Startups

Apart recently published a new blog looking at the structural differences and impact of AI Safety non-profits and for-profit AI labs. Here is an excerpt from 'Where we are on for-profit AI safety':

'AI Safety is crucial, yet the vast majority of investment in AI is funneled into advancing capabilities rather than ensuring safety. Big Labs like OpenAI, once devoted to safety, now prioritize speed and commercialization.’

‘While Big Tech pushes forward with a “move fast and break things” mentality, relatively unsustainable and underfunded non-profits struggle to fill the gap.’

Apart has been (here is a post by Esben, our Co-Founder, on the same topic) and will continue to do a lot of thinking about AI Safety startups over the coming year or so - keep an eye out!

Esben Talks to AI Safety Asia

Esben was invited to speak at AI Safety Asia‘s inaugural roundtable on South-East Asian technical AI safety policy. Esben is a Strategic Advisor for AISA on AI Safety and governance.

He was joined by a really stellar cast: Robert Trager, Co-Director of the Oxford Martin AI Governance Initiative; the former Indonesian Minister, Gita Wirjawan who also joined; and the President of KORIKA, Hammam Riza.

Thanks to Edward Tsoi and Lyantoniette Chua, AISI’s Co-Founders for inviting Esben. AISA is a ‘non-profit organization dedicated to AI Safety and Governance talent development and upskilling across Asia.’

They intend to ‘reduce catastrophic AI risks by identifying and closing gaps on policy levers and transnational coordination for safe and governed AI, starting in Southeast Asia.’

Apart in India

Our Community Program Manager, Archana Vaidheeswaran, was invited by software company Saama to present at their event in Chennai, India. Archana spoke about navigating a path to beneficial AI by building a global community of focused researchers.

Archana’s been building our research collective in India, Singapore, London, Berkeley and across the world. A Singapore 100 Women in Tech 2023 & Board Member at Women in Machine Learning - our community really is in the best hands!

Hacking

AI Safety Startups Hackathon

We asked researchers, engineers and students to think about - over an *intense* weekend - how entrepreneurship might solve the following question: 'Can AI Safety research scale into impactful for-profits?'

Congratulations to the winning ideas!

1st. DarkForest, Mustafa Yasir

2nd. Simulation Operators, Ardy Haroen

3rd. The AI Safety Collective, Lye Jia Jun, Dhruba Patra, Philipp Blandfort

4th. Identity Systems for AIs, Artem Grigor

Google DeepMind's Concordia Contest

This week researchers completed our Concordia Contest: aiming to advance the cooperative intelligence of LLM agents. As AI agents become more capable, understanding how they cooperate is essential as they become able to navigate complex social scenarios.

In cooperation with Google DeepMind's Concordia and Cooperative AI, our participants heard from world-leading experts Sasha Vezhnevets, Joel Leibo, Rakshit Trivedi, and Lewis Hammond.

We saw some really incredible early stage experiments. The contest winners are announced next week!

The Weeks Ahead

• Keep an eye out for Apart's 4th October AI Agent Security Hackathon: 'how do we ensure a successful reevaluation of existing safety frameworks and the development of novel approaches to security in light of AI agents?'
• A longer-form post on the winners of the AI Safety Startup Hackathon.
• Chandler Smith of the Cooperative AI Foundation spoke about the NeurIPS Concordia Contest - hosted in collaboration with Google DeepMind, MIT, UC Berkeley, and UCL. There's a total of $20,000 in prizes: $10,000 for top performers and $10,000 in travel grants for underrepresented groups.

Have a great week and let's keep working towards beneficial and safe AI.

'Apart is a global collective conducting impactful AI Safety research - a decentralized laboratory mobilizing researchers.'