AI for Human Empowerment

​How can we build AI that empowers humans - rather than controls them?

​Join the Effective Altruism Uni Group for a thought-provoking talk exploring a new direction in AI safety. Instead of trying to predict or override human desires, this approach puts human power and autonomy at the center. It’s not about making AI more intelligent - it’s about making it more corrigible, fair, and aligned with human dignity.

​Whether you're curious about AI, concerned about the future, or passionate about ethics and social impact, this talk is for you. Open to all students and young professionals - no prior experience with AI needed!

​Speaker: Jobst Heitzig, Senior Researcher at the Potsdam Institute for Climate Impact Research, brings his experience tackling global challenges - like climate change - to the urgent issue of AI safety. His latest work focuses on creating AI systems that distribute power in safe and just ways, helping humans retain control and make meaningful choices.

​Abstract: Dr. Heitzig's novel research proposes AI systems that support human goals without trying to predict or prioritize them. Instead, these systems fairly manage the distribution of power among individuals, preserving everyone’s ability to shape their own future. This leads to corrigible AI - systems that welcome human oversight and correction, rather than resisting it.

​Come and explore how a non-power-seeking AI could make the world safer for all. We'll bring the snacks, you bring the passion!