Kurzweil tackles one of the biggest scientific mysteries our our ages: How to build a synthetic mind. Instead of descending into mumbo jumbo, touchy feely nonsense, he sets forth a vision that makes a lot of sense, even to sophisticated practitioners. Sure, I think he over simplified the solution, but he gives enough specific hypothesis that one could spend decades fleshing out, refuting, refining his ideas.
All of it
Excellent, extremely pleasant to listen to his voice. Slightly arrogant which suits the material in the book.
Well, the only downside is that Lane mispronounced a few words... especially "von Neumann" repeatedly (saying "von New--man" which is simply not correct. This really bugged me. Clearly Kurzweil did not review the entire performance before giving his ok.
Otherwise his performance was great.
Old & fat, but strong; American, Chinese, & Indian (sort of); Ph.D. in C.S.; strategy, economics & stability theory; trees & machining.
I’m a fan of Kurtzweil. I find his book The Singularity is Near to be an essential part of any modern education. He is also the most financially successful Artificial Inelegance (AI) researcher to date. So he’s a great philosopher, a great researcher, and a great businessmen, but apparently he’s not infallible, because this book missed the mark.
In a nutshell his answer to the question, “How do you/we make a mind” is “The same way we’ve been trying for the last 30 years”. This answer is so close to brilliant it’s spooky. I think there have been profound changes in AI over the last 5 or 6 years. Big data is suddenly revealing (or at least strongly suggesting) that we in fact may have been making pretty good synthetic minds for decades and just didn’t know it. The problem has been that having made a good synthetic mind we don’t know how to educate it. At the same time (because of advances in AI) we may be on the cusp of discovering that we don’t know how to educate biological minds either (and have been making nearly the same mistakes as the AI community has been making). In this book Kurzweil presents much of the data in support of this argument, and then walks away from what to me seemed the logical conclusion.
To make matters worse his overview of 30 years of AI is quite narrow. One might get the impression that he believes that Multi-Layer Hidden Markov Models are all that is needed. Perhaps in some since any of a dozen AI methods are sufficient, in the same since that Truing Machines are sufficient, but the field of synthetic mind creation is much richness than presented here, as the author must have known.
The author lives in a fantasy land. He thinks that he will be able to transfer his mind into a machine.
Avid audiobook addict!
I'm interested in the concept of Artificial Intelligence and thought this book might be interesting because of the author having a name that I'd heard many times in reference to AI, predictions of the future, etc. Instead, it was incredibly boring as the author drones on and on about computer history and inserts a whole bunch of plugs for his various companies.