How to develop AI safely (approaches in AI alignment)

Artificial intelligence makes our lives easier in many ways and offers huge potential for the future, but it also poses one of the greatest risks to humanity. How can it threaten us and what are the latest approaches and considerations in the field of AI Safety?

The evening will start with a short introduction and continue with a panel discussion with AI researchers:
– Walter Laurito – Team Lead & Research Engineer at Cadenza Labs, working on on AI Safety interpretability research
– Claire Short – AI alignment researcher from Chicago, Founder of Athena - a mentorship program for women in AI safety research, Foresight Neurotech Fellow, work with human/model inductive bias, taking inspiration from biological systems to drive hypotheses to research questions
– Benjamin Sturgeon – Researcher from Cape Town, South Africa, working on agency in AI systems, specifically analysing reinforcement learning systems with multiple agents and how to measure/ increase agency in our chosen agents.

Introductory presentation and discussion moderation by Hana Kalivodová

Please complete registration for capacity reasons.