I wouldn't call the article "bad", but I do think it loses points for furthering a "feud" which just need not exist. Norvig-style AI is incredibly useful in the

I wouldn't call the article "bad", but I do think it loses points for furthering a "feud" which just need not exist. Norvig-style AI is incredibly useful in the here and now to solve real-world problems. Hofstadter-style AI is driven by a deep desire to exorcize one of the few remaining "ghosts" in our otherwise natural world. It is an attempt to REMOVE the "ether" of thought stuff, not to hide in it. Perhaps a terminology change is in order? I would call what people like Norvig and Hinton and Ng do "machine learning", while I would probably use "artificial sentience" to describe the Hofstadter-Dennet camp, as that seems to more closely capture their ultimate goal. The world needs both approaches, and hopefully the time will come when the gap is bridged.