Is his lecture any good in your opinion? I scrubbed through several and all I got is him gushing vaguely about some actual ML researchers and stumbling over his own technical slides. What am I missing that this is somehow good enough for MIT?