Artificial Intelligence (AI) is becoming influential throughout technological society, from transport to healthcare, civil engineering to personal electronics. But, what kind of intelligence does our AI possess?
AI pioneer Judea Pearl has said that the state of the art is just very sophisticated fitting of curves to data. This is a long way from intelligent machines we would wish to have: Artificial cognitive systems that support human cognition, and extend thinking beyond human limitations -- in short, human-compatible AI. In order to chart a course towards AI with true intelligence, an AI that is relatable and can be asked to understand us, its users, we must study brain and AI together. The talks in this seminar discuss how this can be done and where it will lead, ranging from anticipated empirical breakthroughs in neural bases of cognition, to the limitations and promises of biologically-inspired learning algorithms, and even to futurism.
Schedule (each speaker 20 minutes):
Riitta Salmelin: Towards individual fingerprints of brain function
Jörg Tiedemann: From language technology to artificial consciousness - Where is the intelligence, and what’s language got to do with it?
Jami Pekkanen: What does machine learning have to do with human learning?
Anna-Mari Rusanen: On Adversarial Attacks
Michael Laakasuo: Human moral attitudes towards Mind Upload
Panel Discussion (30–40 minutes): Human language, algorithmic language, and intelligence
The event is free to enter and free of charge.