Smart Systems, Smarter World - Insights into Machine Learning and Beyond

Everyone who has interacted with ChatGPT knows just how powerful artificial intelligence can be. Yet, beneath this apparent prowess lies a massive neural network trained on virtually every piece of information available online. If these networks can capture the probability distribution of the internet, might they also unravel the elusive probabilities that underpin the quantum world? In quantum mechanics, objects can exist in all their allowed states simultaneously, an idea that turns conventional intuition on its head, especially when many such objects interact. This talk explores how machine learning techniques, particularly neural quantum states, can help navigate the perplexing realm of quantum probabilities and offer fresh insights into phenomena that have long defied our understanding.

We live in a time where people can ChatGPT their way through most of their curiosities. A few clicks are all between you and a well-put-together and pretty accurate analysis of our society. Social science, a field established to document, analyze and explain our social trends seems to have become obsolete in terms of public good. But the real story is more nuanced.

In six different studies, we looked at how advice from artificial intelligence—whether it was written or spoken—affects the way people make moral decisions. We compared AI advice to advice from real humans and found something surprising: people were just as likely to be influenced by AI as by other people. This was true across all kinds of situations and moral dilemmas. It seems that many people treat AI advice like a mental shortcut, using it to help them make tough choices without thinking through all the details themselves. These results raise some important questions about how AI might shape the way we make ethical decisions in everyday life.