HHBC - Emperor's New Mind
HatHead Book Club
a review of forward thinking thinking!
The Emperor's New Mind
Roger Penrose - 1990
Oxford University Press
the other day i discovered that i had a book that i actually had purchased in the future. i first read this book in 2010, twenty years after its first publication but wanted to give it a second attempt with a fresh mind as my first attempt was unsatisifying and so sent the book back in time where i had not yet read it with this note to myself:
"I first picked up this book for a couple of reasons:
i) one of the only science books that chooses to describe the emergence of intelligence in any other way than evolution
ii) is written by Roger Penrose, world renowned mathematician and thinker. "
Then I put the book down only a couple of chapters into it - the book was a weakly camoflauged attempt at denying the mechanical nature of existence: that somehow humans are unique and special in some mysterious way; that man is truly the paragon of animals and at 'the top". This is a philosophy which has only served to distract and constrain scientific achievement since the beginnings of science. There is no room for hubris in scientific thinking."
these are such strong words that i need to reassess the matter - did i miss something? when did i start using captial letters? when did i stop? questions to be answered!
perhaps i didn't understand Penrose's proposition at the time - what is all this fuss that he is making about quantum gravity as the source of human consciousness anyway? just voodoo black magic to me at the time.
now i am better studied on quantum theory and recall Penrose's proposition. since quantum mechanics demonstrates that things literally come from nothing, i was intrigued to see how Penrose applied quantum mechanics to neuroscience - where do our thoughts come from? is our consciousness a product of quantum level interactions? very provocative thinking.
but the book is really a defense around how machines cannot and will not ever have a comparable intelligence to humans; not about the nature of human consciousness.
Warning Signs
there is a tremendous amount of cross-referencing to other pages elsewhere in the book. this style of writing is very distracting and as a reader you are continually forced to break your train of thought and hunt around for the reference. i am always wary of this writing technique - it is as though the author is trying to hide the weakness of his/her argument behind unnecessary complexity. this book is over 500 pages long and i am also aways wary of authors who require so much space to present their thinking; it usually indicates a great deal of time defending inconsistencies.
Penrose's Slippery Slope
on second reading and after a couple of chapters, Penrose's book still comes across as a what-ever-it-takes defense of 'only humans can be genuinely intelligent'. note that the italics in the last quote was not emphasized by me but by Penrose, stepping onto the slippery slope as he tries to make his distinction between "intelligence" and "genuine intelligence". Penrose grants machines 'intelligence' but reserves 'genuine intelligence' for humans. you cannot make something true by simply italicizing it. and in a obvious attempt to skirt the slippery slope, Penrose uses these terms quite liberally in his thought experiments without defining their exact meanings until much later in the book. perhaps Penrose is trying to trick the reader in accepting his terms without defining them?
Penrose puts both feet firmly on the slippery slope when he discusses the classification of Artificial Intelligence (AI). these arbitrary classifications of AI surround the respective models being used to develop AI. Penrose grants that some of these types of AI may be achieved but holds that the existing models will never be equivalent to human intelligence. Penrose completely discounts other potential computing paradigms and their application to AI models simply because they do not yet exist. for example, Penrose glosses over quantum computing as an AI model not constrained by current technology barriers by downplaying any future development of this technology. (but on the other hand, Penrose then holds out for quantum discoveries as the source of understanding human consciousness).
Who Out-Godel's Us?
Penrose digs up the Godel's Incompleteness argument - that things can be true but this cannot be formally determined by the system that they are being described in. what is always avoided by thinkers who hold up human's abilities to "godelize" as our intelligent uniqueness is the necessary consideration that there are also things that we cannot 'godelize', that there are also physical limitiations on our biologically provided intelligence and, therefore, we share the same dilemma of the mechanical machines - we cannot know everything and there will always be things that are 'true' yet we will be unable to see their 'truth'.
To Err Is Human
i had to put Penrose's book back down again when he proposed the intelligence double-standard: if a machine makes a mistake, it is not intelligent. we say "to err is human" and acknowledge the unavoidable human condition of learning from our mistakes. Penrose claims that since Deep Blue made an obvious error when playing chess in game 2 against Gary Kasparov, that making mistakes means that something does qualify as intelligent. Gary Kasparov, in a recent rematch with the machines - a laptop this time, not a super computer- made an obvious mistake that cost him the match. does this make Gary Kasparov, the strongest chess player in the history of the game, unintelligent?
Moving Targets
opponents to artifical intelligence simply keep redefining the meaning of 'articifical intelligence' as 'things machines haven't yet done'. will machines ever have human intelligence? of course not - humans are biological and machines are mechanical. Will machines ever be intelligent? depends who you ask. ask my cat chatbot and it will tell you that it is.
i am going to bring this book back with me into the future so to avoid thinking i need to read it again.
Next HatHead Book Club?
there'll be another one for sure but not sure exactly when - the HatHead Book Club is not for the weak of heart or mind. Stay tuned!
a review of forward thinking thinking!
The Emperor's New Mind
Roger Penrose - 1990
Oxford University Press
the other day i discovered that i had a book that i actually had purchased in the future. i first read this book in 2010, twenty years after its first publication but wanted to give it a second attempt with a fresh mind as my first attempt was unsatisifying and so sent the book back in time where i had not yet read it with this note to myself:
"I first picked up this book for a couple of reasons:
i) one of the only science books that chooses to describe the emergence of intelligence in any other way than evolution
ii) is written by Roger Penrose, world renowned mathematician and thinker. "
Then I put the book down only a couple of chapters into it - the book was a weakly camoflauged attempt at denying the mechanical nature of existence: that somehow humans are unique and special in some mysterious way; that man is truly the paragon of animals and at 'the top". This is a philosophy which has only served to distract and constrain scientific achievement since the beginnings of science. There is no room for hubris in scientific thinking."
these are such strong words that i need to reassess the matter - did i miss something? when did i start using captial letters? when did i stop? questions to be answered!
perhaps i didn't understand Penrose's proposition at the time - what is all this fuss that he is making about quantum gravity as the source of human consciousness anyway? just voodoo black magic to me at the time.
now i am better studied on quantum theory and recall Penrose's proposition. since quantum mechanics demonstrates that things literally come from nothing, i was intrigued to see how Penrose applied quantum mechanics to neuroscience - where do our thoughts come from? is our consciousness a product of quantum level interactions? very provocative thinking.
but the book is really a defense around how machines cannot and will not ever have a comparable intelligence to humans; not about the nature of human consciousness.
Warning Signs
there is a tremendous amount of cross-referencing to other pages elsewhere in the book. this style of writing is very distracting and as a reader you are continually forced to break your train of thought and hunt around for the reference. i am always wary of this writing technique - it is as though the author is trying to hide the weakness of his/her argument behind unnecessary complexity. this book is over 500 pages long and i am also aways wary of authors who require so much space to present their thinking; it usually indicates a great deal of time defending inconsistencies.
Penrose's Slippery Slope
on second reading and after a couple of chapters, Penrose's book still comes across as a what-ever-it-takes defense of 'only humans can be genuinely intelligent'. note that the italics in the last quote was not emphasized by me but by Penrose, stepping onto the slippery slope as he tries to make his distinction between "intelligence" and "genuine intelligence". Penrose grants machines 'intelligence' but reserves 'genuine intelligence' for humans. you cannot make something true by simply italicizing it. and in a obvious attempt to skirt the slippery slope, Penrose uses these terms quite liberally in his thought experiments without defining their exact meanings until much later in the book. perhaps Penrose is trying to trick the reader in accepting his terms without defining them?
Penrose puts both feet firmly on the slippery slope when he discusses the classification of Artificial Intelligence (AI). these arbitrary classifications of AI surround the respective models being used to develop AI. Penrose grants that some of these types of AI may be achieved but holds that the existing models will never be equivalent to human intelligence. Penrose completely discounts other potential computing paradigms and their application to AI models simply because they do not yet exist. for example, Penrose glosses over quantum computing as an AI model not constrained by current technology barriers by downplaying any future development of this technology. (but on the other hand, Penrose then holds out for quantum discoveries as the source of understanding human consciousness).
Who Out-Godel's Us?
Penrose digs up the Godel's Incompleteness argument - that things can be true but this cannot be formally determined by the system that they are being described in. what is always avoided by thinkers who hold up human's abilities to "godelize" as our intelligent uniqueness is the necessary consideration that there are also things that we cannot 'godelize', that there are also physical limitiations on our biologically provided intelligence and, therefore, we share the same dilemma of the mechanical machines - we cannot know everything and there will always be things that are 'true' yet we will be unable to see their 'truth'.
To Err Is Human
i had to put Penrose's book back down again when he proposed the intelligence double-standard: if a machine makes a mistake, it is not intelligent. we say "to err is human" and acknowledge the unavoidable human condition of learning from our mistakes. Penrose claims that since Deep Blue made an obvious error when playing chess in game 2 against Gary Kasparov, that making mistakes means that something does qualify as intelligent. Gary Kasparov, in a recent rematch with the machines - a laptop this time, not a super computer- made an obvious mistake that cost him the match. does this make Gary Kasparov, the strongest chess player in the history of the game, unintelligent?
Moving Targets
opponents to artifical intelligence simply keep redefining the meaning of 'articifical intelligence' as 'things machines haven't yet done'. will machines ever have human intelligence? of course not - humans are biological and machines are mechanical. Will machines ever be intelligent? depends who you ask. ask my cat chatbot and it will tell you that it is.
i am going to bring this book back with me into the future so to avoid thinking i need to read it again.
Next HatHead Book Club?
there'll be another one for sure but not sure exactly when - the HatHead Book Club is not for the weak of heart or mind. Stay tuned!
posted by HatHead at 10:05 AM
4 Comments:
looks like these book clubs are off to a good start!
looks like these book clubs are off to a good start!
the books are from this time. in the future there are no books and all knowledge is transmitted directly into your brain (though spam is still a problem).
depends on the interface (naturally).
Post a Comment
<< Home