I just listened to a great video discussion – Khan Academy and Stanford AI Class: Reinventing Education – with Peter Norvig, Sebastian Thrun, and Sal Khan. It’s a candid discussion of what each of these educators wanted to achieve with opening up their courses and content and some of the challenges they faced in the process. Most importantly, they (particularly Sebastian) discuss where they were wrong in their previous assumptions about learning.

Read the full article here:
Khan and AI: Open Online Courses

Interested as I am in all this, my work is not in education, it is in training. There is, of course, an overlap. Both aim to impart knowledge and foster cognitive skills. But in a training context this is very rarely the endgame. The goal is typically to develop competence, the ability to do a job. And while you can do a reasonable job of measuring knowledge and some cognitive skills using a computer-gradable, online assessment, the majority of competences require a human eye. Of course, graduates of Norvig and Thrun’s course may one day be able to use sophisticated artificial intelligence to assess skills like interviewing, graphic design, welding or using Photoshop, but for now the only reliable judge is another human.

Read the full article here:
Massively scalable training