The Brain and Mind discussion

93 views
Neuroscience > 'How to Create a Mind: The Secret of Human Thought Revealed' by Ray Kurzweil

Comments (showing 1-7 of 7) (7 new)    post a comment »
dateDown arrow    newest »

message 1: by Aaron (new)

Aaron Thibeault (thebookreporter) | 16 comments Just finished reading Ray Kurzweil's new book How to Create a Mind: The Secret of Human Thought Revealed. The book charts the progress of artificial intelligence, and points up how these machines are already operating according to the same principles as our brains. As our understanding of the brain continues to advance, Kurzweil argues, we will be able to use this knowledge to build machines that are ever more human-like--until, eventually, our machines will achieve human-level intelligence, and beyond. The book is inspiring and daunting all at once. I've written an executive summary of the book available here: http://newbooksinbrief.com/2012/11/27...

Cheers,
Aaron


message 2: by Morgan (new)

Morgan Blackledge (morganblackledge) | 4 comments I have to admit. I loved this book. Some say Kurzweil is a crank. I think he's a lot of fun. And I thought this book was really thought provoking at times.


message 3: by Michael (last edited Apr 07, 2013 11:50AM) (new)

Michael (semanticwarrior) | 3 comments Kurzweil is not a crank, he is arguably brilliant, but that still does not mean his predictions are correct. AI will continue to provide more accomplishments in the simulation of intelligent behaviour, but that is a very different type of progress from what Kurzweil claims. I have not read this book but did dip into "The Singularity is Near." I found it to be rather repetitive, using Moore's Law as the main evidence for a series of fascinating but largely unsubstantiated claims. If we focus just on his one claim of software writing software (central to his bootstrapping hypothesis) we see a very large, and in my opinion, insurmountable obstacle. As a software engineer who has read widely on (and used many of) programming languages and the mathematics and logic underlying them, I have experienced first-hand the limitations of these languages. There has been no discernible progress on the fundamental limits of formal systems. If you do some research on Godel, Turing (his mathematics) and Knuth you will agree, I think, that although the tools used to make programs more reliable (compilers, debuggers, and smart editors embedded in sophisticated frameworks such as Spring) have become more feature-rich, faster, and ubiquitous, the theory of (computer) language construction and program proofing still cannot generally assess the veracity of a program in relation to the requirements for the program. And this is true even when the requirements are written (by humans) in very restrictive and precise formats. The jump to a system that can listen to natural language descriptions of what a program should do and produce that program (nevermind a system that will produce the description for another system to consume) is nowhere in sight. This task is still the hardest problem in computer science! Despite the production of hundreds of millions of computer programs, the large majority of systems are over-budget, delivered late, do not meet the requirements as judged by the people commissioning the effort, and most importantly, are bug-rich. The singularity, as Kurzweil describes it, will not happen in 2029, nor in the lifetime of anyone now alive. I don't believe it will ever happen. Now if someone wants to discuss ideas of how we (as a species) might use computer tools to design variations on biologic frameworks (genetic, proteomic, metabonomic, etc) leading to products of artificial evolution, let's have at it :)! Some influential reading on my part in this arena has been Oryx and Crake (Margaret Atwood, a fictional treatment) and The Birth of Mind (Gary Marcus, exemplary science writing, by a real scientist).


message 4: by George (new)

George A. (gsimpso4) | 2 comments I am a fan of Kurzweil, but this book does not achieve what it claims. A brain and a mind are not the same thing. The difference is that a mind is something that interacts with the conceptual field - the vast body of ideas, information,and concepts that humans uniquely have access to, among the creatures on this planet.

I have recently published my first book on the same theme, which I think does achieve the goal Kurzweil sets out.


message 5: by Oné (new)

Oné (Baldscientist) I do not dislike Kurzweil, but I do not particularly like his ideas on neuroscience. He's brilliant, no question, but he is a little out of his league in the matter of the mind. The thing is that he and most people that look at the problem from an engineering point of view. Nothing wrong with that, except that almost without exception, this point of,view is based purely on the brain's connections and the modeling of such connections. There are many, and I mean MANY other components of brain function besides the physical connections. By that I mean neurotransmitters that "spill" to neurons that are not directly connected, the presence and physiology of more than one transmitter in the synapse, glial-neuron interaction and quite a few more. I may be mistaken, but I do not think the "connections" only school of thought take into consideration these type of factors...


message 6: by Michael (new)

Michael (semanticwarrior) | 3 comments @George. When you speak of the 'conceptual field' it brings to mind the language of the Gestalt School of Psychology. Do you really mean to imply that there is something more to the brain than the matter it is composed of?

@One. To follow on my first remark here, the mind is definitely more than the connections. I agree with your observations. Whatever allows us the ability of the mind is certainly not going to be captured by a 'network model.' Hence the limitations of Kurzweil vision, and the science of AI in general. Not to say it isn't extremely useful...but it is not intelligence.


message 7: by George (new)

George A. (gsimpso4) | 2 comments Sorry for the delay in responding. My book "Conscious Patterns Dancing in the Field: the Physics of Mind" argues that consciousness actually is a pattern, a computational pattern, a pattern that dances across the brain’s surface as we think.

The thing that distinguishes human consciousness from that of animals is that humans' patterns act on the interface between the abstract reality of ideas and concepts to the physical world; they link these two aspects of reality.

As far as I know, this idea is new. I like it because it is supported by evidence, it is testable when the equipment becomes capable enough, and it has significant practical implications.


back to top