Constellation’s mission is to help the world safely navigate the development of transformative AI.

Our goals

Our goals

In alignment with our mission, we have two primary goals:

1.

Accelerate safety research

We bring together people doing cutting-edge work on the safety of powerful AI systems, enabling them to learn from and collaborate with each other. Many researchers tell us that their research is better or even that their research direction has changed as a result of collaboration at Constellation.

2.

Develop safety talent

Via our fellowships and programs, we support high-potential individuals who want to help the world safely navigate the development of transformative AI. Dozens of participants from our past programs now work full-time on leading safety teams.

The Constellation team

Core team

Programs at Constellation

Programs are held in Berkeley, CA.

Alexander Beck

Executive Chef

Beck created, and leads, the culinary program, maintaining wide-ranging menus that provide a good basis for Constellation members and visitors to get to know each other. In addition to AI, he is also interested in data science and politics. Within politics, his interests include campaign management, poll analysis and direct persuasion.

Alex Fields

Senior Program Manager

Alex works on revising and expanding Constellation’s programs. Previously, he was chief of staff at the Atlas Fellowship. Prior to that, he was a management consultant.

David Espiritu

Head of Business Operations

David’s remit includes financial planning, accounting, HR and compliance. In addition, he advises on IT and works on special projects. He has a background in finance, and he studied mathematics and economics at University of California, Davis. Alongside his work at Constellation, David is a math tutor and mental health coach.

James Bregan

CEO & Co-Founder

James followed the AI safety field since the mid 2000's before helping build Constellation. His career has focused on startups and entrepreneurship in both the non-profit and for-profit sectors. He was a key technology leader at PayPal as the company scaled from 100 to 10,000 people, and served as EVP Engineering during its acquisition by eBay.

Keith Weissglass

Director of Operations

Keith leads the Program Operations team, which manages Constellation’s research center and supports its AI safety programs. Keith’s previous work includes management consulting at Gartner and leading product marketing at several successful tech startups.

Lauren Mangla

Program Manager

Lauren manages Constellation’s Visiting Fellows program and other AI safety field-building initiatives. Previously, she was the executive director of SPAR, an AI safety upskilling program. Lauren also has experience in policy and scientific research.

Linh Nguyen

Recruiting Operations

Linh manages our hiring pipeline and refines our recruiting processes. Before working at Constellation, Linh spent ten years at a worker-owned collective. There, she learned about democratic decision-making, baked bread, and traveled to Europe to source cheeses.

Lucy Mahaffey

Program Operations

Lucy spearheads operations at Constellation, especially around events, programs, and membership requests. Prior to Constellation, she coordinated transatlantic initiatives between US and UK legislators. Lucy’s previous experience includes ten years of research into human trafficking. She has degrees in international relations and sociology from OU, Nottingham, and Cambridge.

Advisors

Programs at Constellation

Programs are held in Berkeley, CA.

Bill Zito

Advisor

Bill is a co-founder and former President of Constellation. Bill previously co-founded and helped run Redwood Research. At Aurora, a now-public autonomous vehicle startup, Bill was the first engineer working on ML labeling pipelines. He helped build three engineering and operations teams around that work as the company scaled from 30 to 300.

Nate Thomas

Advisor

Nate is a co-founder and former CEO of Constellation. Nate previously co-founded and led Redwood Research, where he led a team researching AI control, interpretability, and other safety-relevant areas. Nate has published work on AI safety, condensed matter physics, machine learning for biomolecular structure prediction, and quantum gravity. He has degrees in mathematics, physics, and computer science from MIT, Cambridge, Oxford, and Stanford.