AlignmentScientist/Engineer

There are more good alignment ideas than there are people to run them. Top researchers at Anthropic, DeepMind, OpenAI, and the major alignment organizations all have promising directions they can’t get to, ideas that could be more impactful than what’s currently getting funded or what their organizations are set up to try. The window for alignment research is short, and we think the field’s real bottleneck is capacity, not ideas. We’ve built something to fix that. AE Studio is a 160-person, fully bootstrapped ML consultancy. We’ve spent over a decade building and shipping AI systems for clients including Samsung, Berkshire Hathaway, Goodfire AI, and RAND Corporation. Without outside investors we are able to choose where the money goes, and we put it into alignment research through the AI Alignment Foundation, a nonprofit we founded to scale this work (also backed by other philanthropies). The research team is 25 people and growing. Over years of client work, we got really good at one thing: taking someone with a vision (an external visionary) and no capacity to build it, putting a strong technical team around them, and shipping fast. Our alignment research is that same approach pointed at the field’s best ideas instead of products. The visionary doesn’t hire anyone, manage anyone, or babysit anyone. They plug in, have a conversation once a week, and our team runs the work. We’re running a DARPA seedling on alignment and control, our policy team engages directly with the government on AI alignment, and our work has been featured many times in the Wall Street Journal. We collaborate with researchers at Anthropic, Princeton University, University College London, Redwood Research, and MATS. The work gets published at major ML conferences and workshops. ABOUT THE ROLE You will be part of the alignment research team, working on projects selected for their high upside potential and under-resourced status. Research that matters most in a world where AI systems are powerful