“What’s next for Artificial Intelligence?”

What can we expect from Artificial Intelligence? Four experts share their vision of expected advances in algorithms that are capable of improving their programming on their own. Yann LeCun (Facebook) wonders about the most effective way to teach machines to self-program and believes that “deep learning” will take us further than “machine learning.” Andrew Ng (Baidu) analyses the potential effects of AI on work organisation and predicts a major shake-up in employment that is “unprecedented since the 1930s.” Nick Bostrom (Future of Humanity Institute in Oxford) ponders the need to control AI to prevent its freewheeling algorithms from posing a threat to humankind.
Luke Nosek (PayPal) compares the promises and dangers of AI to those of nuclear energy, and concludes that humans need to train their brains before the artificial ones arrive.