Introduction to AI Safety. Let's be ready before it's too late!
Sunday, May 28
15:10 - 15:40
What do you think AI will be capable of in 10 years? And in 20 or more? Should we start to be worried now and be prepared to prevent malicious behaviors? Let’s discover a new discipline called AI Safety that addresses the potential risks of AI and how to prevent them.
Thanks to the power of neural networks (Deep Learning) and increasingly powerful tools, what we can now come close to call Artificial Intelligence is spreading in all kinds of disciplines from image and video generations to suggestions for potential new scientific discoveries.
Should we begin to worry about such power? Are our human moral values aligned with nascent AIs? Even the global Effective Altruism community recognizes the risks of Artificial Intelligence as one of the existential risks for humanity!
Fortunately, humans have already started to recognize it. Under a new discipline called AI Safety, currently several research centers and universities are working to identify major problems and research questions to collaborate together toward the best preventive solutions.
The talk will include some examples of work in Python from international curricula.
Stefania Delprete
Stefania Delprete gestisce il capitolo italiano di Effective Altruism (Altruismo Efficace) e un gruppo professionale di persone esperte o aspiranti nel campo della data science, machine learning ed intelligenza artificiale coinvolte in tale comunità di altruisti efficaci.
Studia fisica in Italia per poi lavorare nella programmazione nel Regno Unito e in Germania. Negli ultimi anni di è occupata di progetti e insegnamenti di Python e data science.