Eliezer Yudkowsky, who coined the term, argues that developing helpful AI needs to be a higher research priority: it might require a large expense and it should be concluded just before AI turns into an existential https://bit.ly/AIInVideo