Why do you care about AI Existential Safety?

We have lost control of the ability to predict all usable technology behavior. Artificial intelligence continues to develop at a dizzying pace and enter more and more into our lives. This process is inevitable. At some point in the future, artificial intelligence may aspire in directions that do not align with humanity’s desires. We need to prepare for this moment and prepare tools that will help us monitor such deviations.

If you have got here, please read Isaac Asimov’s wonderful short story: All the Troubles of the World:
https://schools.ednet.ns.ca/avrsb/070/rsbennett/HORTON/shortstories/All%20the%20troubles%20of%20the%20world.pdf

Please give at least one example of your research interests related to AI existential safety:

My latest published research: Clever Hans or Neural Theory of Mind? Stress Testing Social Reasoning in Large Language Models
Following the escalating debate on AI’s theory of mind capabilities, I suppose you know the wolf story, I found it important to provide an objective picture of what we currently have.
https://aclanthology.org/2024.eacl-long.138.pdf

I am now continuing my research in developing novel methods for identifying theory of mind abilities and also hope to find ways to control those skills.