What do you think?
Rate this book
The Most Human Human: What Talking with Computers Teaches Us About What It Means to Be Alive
Goodreads Choice Award
Nominee for Best Nonfiction (2011)The Most Human Human is a provocative, exuberant, and profound exploration of the ways in which computers are reshaping our ideas of what it means to be human. Its starting point is the annual Turing Test, which pits artificial intelligence programs against people to determine if computers can “think.”
Named for computer pioneer Alan Turing, the Turing Test convenes a panel of judges who pose questions—ranging anywhere from celebrity gossip to moral conundrums—to hidden contestants in an attempt to discern which is human and which is a computer. The machine that most often fools the panel wins the Most Human Computer Award. But there is also a prize, bizarre and intriguing, for the Most Human Human.
In 2008, the top AI program came short of passing the Turing Test by just one astonishing vote. In 2009, Brian Christian was chosen to participate, and he set out to make sure Homo sapiens would prevail.
The author’s quest to be deemed more human than a computer opens a window onto our own nature. Interweaving modern phenomena like customer service “chatbots” and men using programmed dialogue to pick up women in bars with insights from fields as diverse as chess, psychiatry, and the law, Brian Christian examines the philosophical, biological, and moral issues raised by the Turing Test.
One central definition of human has been “a being that could reason.” If computers can reason, what does that mean for the special place we reserve for humanity?
Named for computer pioneer Alan Turing, the Turing Test convenes a panel of judges who pose questions—ranging anywhere from celebrity gossip to moral conundrums—to hidden contestants in an attempt to discern which is human and which is a computer. The machine that most often fools the panel wins the Most Human Computer Award. But there is also a prize, bizarre and intriguing, for the Most Human Human.
In 2008, the top AI program came short of passing the Turing Test by just one astonishing vote. In 2009, Brian Christian was chosen to participate, and he set out to make sure Homo sapiens would prevail.
The author’s quest to be deemed more human than a computer opens a window onto our own nature. Interweaving modern phenomena like customer service “chatbots” and men using programmed dialogue to pick up women in bars with insights from fields as diverse as chess, psychiatry, and the law, Brian Christian examines the philosophical, biological, and moral issues raised by the Turing Test.
One central definition of human has been “a being that could reason.” If computers can reason, what does that mean for the special place we reserve for humanity?
303 pages, Hardcover
First published March 1, 2011
About the author
Brian Christian
10 books831 followersBrian Christian is the author of The Most Human Human, which was named a Wall Street Journal bestseller, a New York Times Editors’ Choice, and a New Yorker favorite book of the year. He is the author, with Tom Griffiths, of Algorithms to Live By, a #1 Audible bestseller, Amazon best science book of the year and MIT Technology Review best book of the year.
Christian’s writing has been translated into nineteen languages, and has appeared in The New Yorker, The Atlantic, Wired, The Wall Street Journal, The Guardian, The Paris Review, and in scientific journals such as Cognitive Science. Christian has been featured on The Daily Show with Jon Stewart, Radiolab, and The Charlie Rose Show, and has lectured at Google, Facebook, Microsoft, the Santa Fe Institute, and the London School of Economics. His work has won several awards, including fellowships at Yaddo and the MacDowell Colony, publication in Best American Science & Nature Writing, and an award from the Academy of American Poets.
Born in Wilmington, Delaware, Christian holds degrees in philosophy, computer science, and poetry from Brown University and the University of Washington. A Visiting Scholar at the University of California, Berkeley, the Director of Technology at McSweeney’s Publishing, and an active open-source contributor to projects such as Ruby on Rails, he lives in San Francisco.
Christian’s writing has been translated into nineteen languages, and has appeared in The New Yorker, The Atlantic, Wired, The Wall Street Journal, The Guardian, The Paris Review, and in scientific journals such as Cognitive Science. Christian has been featured on The Daily Show with Jon Stewart, Radiolab, and The Charlie Rose Show, and has lectured at Google, Facebook, Microsoft, the Santa Fe Institute, and the London School of Economics. His work has won several awards, including fellowships at Yaddo and the MacDowell Colony, publication in Best American Science & Nature Writing, and an award from the Academy of American Poets.
Born in Wilmington, Delaware, Christian holds degrees in philosophy, computer science, and poetry from Brown University and the University of Washington. A Visiting Scholar at the University of California, Berkeley, the Director of Technology at McSweeney’s Publishing, and an active open-source contributor to projects such as Ruby on Rails, he lives in San Francisco.
Ratings & Reviews
Friends & Following
Create a free account to discover what your friends think of this book!
Community Reviews
Displaying 1 - 30 of 478 reviews
July 9, 2014
This is one of the most poorly written must-reads I've ever read. The topic of artificial intelligence is very important and well-researched by the author. But the delivery is couched in half-baked philosophy and capital abuse of the poetic license. Many times I paused the audiobook to yell at Brian Christian.
"Just say what you MEAN!"
"I don't think that word means what you think it means!"
This book's saving grace is that he did a really cool thing. He was a human "confederate" on the Loebner prize. This contest implements the famous Turing Test. There's a prize for the most human computer, and the most human human, thus the book's title. The author won the later prize, mostly by typing everything that popped into his head, but also by studying what makes people vote "human" on the other end.
This journey takes us through all the different tricks AI programmers use to make their chat bots believable. It doesn't go into depth on algorithms, but it doesn't really need to. When he tries to talk jargon, he gets it horribly wrong. I remember something about "statistical" vs. "algorithmic" programming that made no sense to this reader with a bachelor's degree in computer science...
The book also talks about the history of AI programming, from checkers to chess, etc.
One of the stupidest parts of the book was the theme early on with the left vs. right hemisphere of the brain. Christian dragged this neuroscience fact through the dirt and into the realm of nonsense philosophy. He argues that we should teach "right-brained" stuff in school like dance. This is wrong. You don't need a PhD in dance to be a dancer. You do need a PhD physics to be a doctor or physicist. These disciplines are different, and one has unique value above the other simply because it's useful to society. Liberal arts degrees do grow on trees, you know.
He also thinks that Catholic guild, Von Neumann architecture, and analytical thinking are all left-brained. They aren't. That's a pernicious and foolish myth. Brian Christian should know better.
He has this weird fertilization of "analytical thinking" that it's somehow lower than other types of thinking simply because it's how computers think. This is backwards. It's the highest form of thinking, and that's why we're able to program it into computers. Other forms of thinking are more messy, less objective, and less specific to the real world. Other forms of thinking may be site-specific (to use one of Brian Christian's favorite pieces of jargon) to your situation or your own feelings, but that doesn't make these other forms of thinking better in any way.
Take away all that personal context, and the only type of thinking you have left at the bottom of the barrel is a thin film of "analytical thinking".
I sound like Brian Freaking Christian right now...
"Just say what you MEAN!"
"I don't think that word means what you think it means!"
This book's saving grace is that he did a really cool thing. He was a human "confederate" on the Loebner prize. This contest implements the famous Turing Test. There's a prize for the most human computer, and the most human human, thus the book's title. The author won the later prize, mostly by typing everything that popped into his head, but also by studying what makes people vote "human" on the other end.
This journey takes us through all the different tricks AI programmers use to make their chat bots believable. It doesn't go into depth on algorithms, but it doesn't really need to. When he tries to talk jargon, he gets it horribly wrong. I remember something about "statistical" vs. "algorithmic" programming that made no sense to this reader with a bachelor's degree in computer science...
The book also talks about the history of AI programming, from checkers to chess, etc.
One of the stupidest parts of the book was the theme early on with the left vs. right hemisphere of the brain. Christian dragged this neuroscience fact through the dirt and into the realm of nonsense philosophy. He argues that we should teach "right-brained" stuff in school like dance. This is wrong. You don't need a PhD in dance to be a dancer. You do need a PhD physics to be a doctor or physicist. These disciplines are different, and one has unique value above the other simply because it's useful to society. Liberal arts degrees do grow on trees, you know.
He also thinks that Catholic guild, Von Neumann architecture, and analytical thinking are all left-brained. They aren't. That's a pernicious and foolish myth. Brian Christian should know better.
He has this weird fertilization of "analytical thinking" that it's somehow lower than other types of thinking simply because it's how computers think. This is backwards. It's the highest form of thinking, and that's why we're able to program it into computers. Other forms of thinking are more messy, less objective, and less specific to the real world. Other forms of thinking may be site-specific (to use one of Brian Christian's favorite pieces of jargon) to your situation or your own feelings, but that doesn't make these other forms of thinking better in any way.
Take away all that personal context, and the only type of thinking you have left at the bottom of the barrel is a thin film of "analytical thinking".
I sound like Brian Freaking Christian right now...
September 7, 2017
"Algorithms To Live By" was such a special find. It spoke to me and how I think in so many ways. I know human behaviour is weird and varied, but there are patterns. And I am always looking to streamline what I do. And I am always asking why do we do it that way? Surely there's a better way to do this!
So, I've been getting around to Brian's previous book for a few months. And it was also magnificent. Not quite so good, but still a humdinger!
Brian takes us into his world of the Turing Test. Of being a designated human in the lot. He looks at how similar our cultures and behaviour are away from a potential programmed AI, and he talks of how computers were made in our image, and how we make ourselves into their image. There are some meaty ideas in here.
And I listened to it via Audible, read by Brian himself who is a wonderful speaker. His reading is full of nuance and is much more like a lecture than a reading.
So, I've been getting around to Brian's previous book for a few months. And it was also magnificent. Not quite so good, but still a humdinger!
Brian takes us into his world of the Turing Test. Of being a designated human in the lot. He looks at how similar our cultures and behaviour are away from a potential programmed AI, and he talks of how computers were made in our image, and how we make ourselves into their image. There are some meaty ideas in here.
And I listened to it via Audible, read by Brian himself who is a wonderful speaker. His reading is full of nuance and is much more like a lecture than a reading.
August 6, 2019
I have a special interest in the philosophy of the mind - and I love reading and re-reading what people have to say about the brain, or mind, or soul. The computational theory of mind is the main angle that Christian explored in this book through the Turing Test and its implications. An AI passes the Turing Test when it's indistinguishable from a human, usually determined through conversation with another human. And there were times throughout history where different AI programs did pass the test - even if it was just for 5 minutes. Christian presents a wonderful mish-mash of thoughts, comparisons, personal experiences, and research. The book was so stuffed full of ideas and I devoured them eagerly.
Christian's book can be rambling a large portion of the time - it feels like a raw unedited draft with all his ideas and thoughts. It didn't bother me that much because I'm intrigued by most things he's wrote - but it would have been better if it was more structured.
Something he said struck me: that currently, one of the biggest things that separates humans from AIs is that a human actively responds to new information and it's environment. We essentially change after every single new experience we have, because we are always adapting, updating ourselves. Whereas a programmed AI is limited to its program. Even when it's programmed to learn from new experiences it doesn't decide how it learns, only what. (I'm probably not doing this justice and I'm probably wrong, but that was my takeaway.) It blew me away.
"Our very essence is a kind of mongrelism. It strikes me that some of the best and most human emotions come from this lichen state of computer/creature interface, the admixture, the estuary of desire and reason in a system aware enough to apprehend its own limits, and to push at them: curiosity, intrigue, enlightenment, wonder, awe."
Christian's book can be rambling a large portion of the time - it feels like a raw unedited draft with all his ideas and thoughts. It didn't bother me that much because I'm intrigued by most things he's wrote - but it would have been better if it was more structured.
Something he said struck me: that currently, one of the biggest things that separates humans from AIs is that a human actively responds to new information and it's environment. We essentially change after every single new experience we have, because we are always adapting, updating ourselves. Whereas a programmed AI is limited to its program. Even when it's programmed to learn from new experiences it doesn't decide how it learns, only what. (I'm probably not doing this justice and I'm probably wrong, but that was my takeaway.) It blew me away.
August 17, 2016
Um bom livro, mas com pouca coisa nova (para mim). Comecei a lê-lo por ser do mesmo autor do Algorithms to Live By: What Computers Can Teach Us About Solving Human Problems, mas não achei tão bom. O autor participa de um teste de Turing em 2009 e resolve ser o humano mais humano no teste, ou seja, mostrar o que só humanos sabem fazer. E o livro é uma digressão bem legal sobre como nossa mente funciona e como computadores e algoritmos tem facilidades ou dificuldades com o que fazemos. Bastante informação curiosa e legal, mas o A Informação e Como Criar uma Mente: os segredos do pensamento humano já trazem o conteúdo da melhor parte e mais explicado. Bons insights, dá para ver como ele estava caminhando para o Algorithms to Live By: What Computers Can Teach Us About Solving Human Problems, mas este mais recente dá exemplos bem mais concretos e leva isso muito mais além. Vale mais a leitura.
April 21, 2011
Quite simply, the best book I've read, ever. I'm compelled get up out of bed and write down some thoughts after finishing The Most Human Human. I did a double-take when Christian wrote about listening to the Spice Girls in middle school. He writes way beyond his 26 years. On the other-hand maybe those further along in years are writing ever so slightly off the pulse of the intersection of humans and technology. In contrast to What Technology Wants by Kelly, a journalist and also a favorite of mine, Christian passes some sort of authenticity tech guru turing test.
I point out his age because, in essence, what I'm left with is faith that we, humans, get better with each generation just as Christian's generation is showing the world. That despite the unbounded trajectory of computing power, Humans are also unbounded. Computing will keep taking ground but humans will keep taking the higher ground. Christian writes that the year the Turing test is passed is not the year to watch. It's the next year, when Humans roll up their sleeves and prove that we are context aware, anti-cliche, anti-summupable, autoincorrectable, entropy-rich, uncompressable and have awoken once more from complacency. The singularity IS near but humans will continue to lead computing right past that non- event. I have new faith that my sons and their generation will be smart as hell as a result of this sublime tension of humans and computing.
I point out his age because, in essence, what I'm left with is faith that we, humans, get better with each generation just as Christian's generation is showing the world. That despite the unbounded trajectory of computing power, Humans are also unbounded. Computing will keep taking ground but humans will keep taking the higher ground. Christian writes that the year the Turing test is passed is not the year to watch. It's the next year, when Humans roll up their sleeves and prove that we are context aware, anti-cliche, anti-summupable, autoincorrectable, entropy-rich, uncompressable and have awoken once more from complacency. The singularity IS near but humans will continue to lead computing right past that non- event. I have new faith that my sons and their generation will be smart as hell as a result of this sublime tension of humans and computing.
April 17, 2012
This is a fine book. Which is a huge disappointment, because it could have been excellent. It has one of the best premises--and best titles--of any book to come out recently. It got a lot of press, because the interest in the topic is immediate and obvious.
With all that, I wanted a story of the Loebner Prize and the author's quest for the Most Human Human award, along with some computer science and philosophy. I didn't get a story of the Loebner Prize--at all. He talks about leading up to it, makes a few references to the actual event randomly throughout, and then skips to the award ceremony, which was profoundly disappointing.
As for the computer science and philosophy, well... There's a lot of it.
Christian talks about human speech patterns, which, in spoken conversation, are overlapping, interrupting, digressing--everything but linear. Maybe it was his intent to write the book that way, but if it was, it's an impressive failure. There are some chapters that have a coherent direction. But most of them randomly wander off and never get to the point. You might find yourself suddenly reading about Aristophanes in Plato's Symposium and never quite figure out how that has anything at all to do with the Turing test. Sometimes he'll give a bit of a precis at the start of the chapter, but then he won't follow it.
And there are all these headers throughout, which contain no structural value at all. It seems like he wrote the thing and then just stuck a header in whenever a pun occured to him. Or whenever he realized that this page-long bit has nothing to do with anything in the text surrounding it. And they're all A-heads, so they don't help create a sense of hierarchy. There are also random epigraphs, sometimes in the middle of a section. I get it, dude, you studied philosophy and poetry. You know stuff. But you need to work all of this material together into a coherent whole, instead of leaving all the lumps in the gravy.
So as you read, you find yourself in a forty-page long digression about data entropy and compression algorithms. Or not even a digression, really, because "digression" implies some sort of starting off point, and the only relation this seems to have to the purported topic of the book is that he read about it while researching the book. And there are some interesting things about how video compression works in there, but somewhere in the hour or so of reading this chapter, you start to wonder what the hell happened to the narrative.
The chapters that I liked best were the ones I read all in one sitting--it seems like you need to take this book in hundred-page chunks in order for him to wander back to his topic often enough to figure out what's going on.
There are interesting things in this book. But--honestly--the interesting things have already been mined. I've heard several radio stories based on this book, and a few more brushing the same topics (on This American Life and Radiolab). Those stories were a lot better than this book is. Which shows what a really talented journalist can do with the material. Christian, on the other hand, mostly squanders it.
With all that, I wanted a story of the Loebner Prize and the author's quest for the Most Human Human award, along with some computer science and philosophy. I didn't get a story of the Loebner Prize--at all. He talks about leading up to it, makes a few references to the actual event randomly throughout, and then skips to the award ceremony, which was profoundly disappointing.
As for the computer science and philosophy, well... There's a lot of it.
Christian talks about human speech patterns, which, in spoken conversation, are overlapping, interrupting, digressing--everything but linear. Maybe it was his intent to write the book that way, but if it was, it's an impressive failure. There are some chapters that have a coherent direction. But most of them randomly wander off and never get to the point. You might find yourself suddenly reading about Aristophanes in Plato's Symposium and never quite figure out how that has anything at all to do with the Turing test. Sometimes he'll give a bit of a precis at the start of the chapter, but then he won't follow it.
And there are all these headers throughout, which contain no structural value at all. It seems like he wrote the thing and then just stuck a header in whenever a pun occured to him. Or whenever he realized that this page-long bit has nothing to do with anything in the text surrounding it. And they're all A-heads, so they don't help create a sense of hierarchy. There are also random epigraphs, sometimes in the middle of a section. I get it, dude, you studied philosophy and poetry. You know stuff. But you need to work all of this material together into a coherent whole, instead of leaving all the lumps in the gravy.
So as you read, you find yourself in a forty-page long digression about data entropy and compression algorithms. Or not even a digression, really, because "digression" implies some sort of starting off point, and the only relation this seems to have to the purported topic of the book is that he read about it while researching the book. And there are some interesting things about how video compression works in there, but somewhere in the hour or so of reading this chapter, you start to wonder what the hell happened to the narrative.
The chapters that I liked best were the ones I read all in one sitting--it seems like you need to take this book in hundred-page chunks in order for him to wander back to his topic often enough to figure out what's going on.
There are interesting things in this book. But--honestly--the interesting things have already been mined. I've heard several radio stories based on this book, and a few more brushing the same topics (on This American Life and Radiolab). Those stories were a lot better than this book is. Which shows what a really talented journalist can do with the material. Christian, on the other hand, mostly squanders it.
August 8, 2015
In principle, I should have loved this book. I did my PhD in machine learning. I have a fascination with philosophy. But, I found it somewhat tedious at times. I kept wanting to hear more about his actual experience with the competition. He should have included the transcript, or at least more excerpts. But, there was not very much about his experience with the competition at all. Instead it was a collection of examples of how humans differ from computers, or how ideas from AI apply to humans, etc. I thought the last major chapter on compression could have, ironically, been compressed.
On the plus side, it's a very unique book. There were some very interesting observations. It was well written, and at times the writing was quite funny.
On the plus side, it's a very unique book. There were some very interesting observations. It was well written, and at times the writing was quite funny.
September 11, 2012
I was hoping for more of the artificial intelligence part of this book, but it turned out to be more "what we can do better than AIs", which wasn't quite what I was interested in. It's an interesting meditation on what sets us apart, in some places, though it's lacking in organisation -- if I tried to turn in my dissertation with such random chaptering and subtitles, I'd be whacked over the head with the red pen of loving correction by my supervisor. It didn't flow at all well. And I know it's non-fiction, but it felt clunkily info-dumpy. Half the time I was going duh, I know this stuff, that's why I'm reading this book and the other half whoa, slow down.
I think this could be a very interesting book, if it caters to what you're interested in. I was more interested in the artificial intelligences, of which there's very little direct discussion...
I think this could be a very interesting book, if it caters to what you're interested in. I was more interested in the artificial intelligences, of which there's very little direct discussion...
April 12, 2014
At last, proof positive that degrees in philosophy and poetry superbly complement a degree in computer science! TMHH is an extended think piece, thread through with prose poetic writing, on how human minds and silicon ones are alike and how they differ. It is a hugely thought-provoking book, sentence after paragraph after page after page. Judging from its jacket photo, the author is a pretty young guy (a striking contrast with the depth and maturity of the book in review). May he have a long and fruitful writing career ahead of him. And may he continue cogitating on the blurring boundaries between man and machine. There's enormous wisdom in his thinking. Splendid book.
June 21, 2011
First there was Eliza. Then fractal music giving way to database-rich Bach- and Beethoven-simulators. Then Deep Blue. Then Watson. Soon… R. Giskard Reventlov of Aurora? Philip K. Dick’s Preserving Machine? Suffice it to say that The Most Human Human is one of the best nonfiction books it has been my pleasure to read. It touches on all my favorite topics -- recreational math, information theory, philosophy, social psychology, virtual vs. genuine identity – it’s like John Searle meets William Poundstone as might have been channeled through Alex Trebek.
Those wishing to preview the themes and overall plot of the book in greater depth can read the author’s own 8-page online redaction. I did, and can assure you it only whetted my appetite to read the book itself. Nor did this book disappoint.
For those too time-strapped or lazy to check out the redacted version, here’s a thumbnail sketch. If you only had 5 minutes of text-only IM’ing to convince somebody you were YOU, could you do it? How would a stranger figure out that yours were not the automated responses of a preprogrammed ‘bot, nor that of a phisher, nor even that of David Bowie, but YOU. (Umm, uunless, of course, you *are* David Bowie, in which case, may I suggest you use this book’s premise for your next concept album?) Could you convince them? And how would you do it? How do we recognize each other as distinct individuals? How does one measure the soul? Who are we really?
That’s the challenge that Christian sets for himself, and if this strikes you as a bit arbitrary as personal challenges go, then you need a bit of background on Alan Turing’s test and its annual incarnation as the Loebner Prize.
Alan Turing was a math genius who cracked the Nazi’s Enigma code and helped invent the modern-day computer among other things. (This bit of history is fascinating and well worth your time, as you get to meet not only Turing, but guys like Claude Shannon, Johnny Von Neumann, John Nash, and others. Fiction readers can find them in Neal Stephenson’s Cryptonomicon and the film “A Beautiful Mind;” casual nonfiction readers will find them in books by William Poundstone, James Gleick, and Richard Rhodes, and all of these I highly recommend.
Turing was fascinated by the idea of artificial intelligence and thinking machines and proposed that the only way to tell the difference between a sentient and nonsentient being was through a basic conservatory audition. Only, instead of some aspiring floutist piping their way through Fauré’s Morceau des Concours in hopes of an appointment, it was an invisible conversant. If the auditors believed their interlocutor to be human, then voila! Whatever turned out to be behind the screen would have to be considered capable of independent thought.
Seem crazy? Yeah, crazy like a fox, and it was only a matter of time before the Turing test got institutionalized in the form of the Loebner Prize competition, that 5-minute IM thing to which Christian wangled himself an invitation (thus the challenge at the core of this book). However, Christian isn’t content to “just be himself” in the hope and expectation that so doing will prove sufficient to win over the Loebner’s panel of PhD’d judges (psychologists, linguists, computer programmers, etc.), nor does Christian want merely to outcompete the various chatbots entered alongside him and his fellow human controls/confederates. No, Christian is out to prove to all and sundry that not only is he capable of demonstrating his humanity anonymously to strangers, he can do so better than any other human in the competition. Much of the book deals with his efforts as a kind of Rocky Balboa IV, training himself to challenge the Ivan Dragos of chatbots and fellow panelists to a human-off.
Think it’s easy, do you? Try sitting down next to someone at random on the Acela and striking up a conversation without mentioning the weather or running some other personally pre-programmed bit of dialogue: “Hi, how are you, my name’s ___, where are you headed, etc.” Unless you’re a transactional analyst, you’d be surprised how much of our lives come to us pre-scripted in your typical five-minute span.
The issues touched on here run from the sublimity of learning to live life to its fullest (by avoiding rote, repetitive experiences) to the ridiculous informational density of language (he shares his results playing the “Shannon Game” at p. 227 -– I tried it and can tell you that the phrases given are bizarre, and not always grammatical; anyway, it’s worth playing once). Along the way, he offers a rogues gallery of chatbots -- from the ever-popular therapist Eliza and her patient foil Manny all the way to present-day AIs like the U.S. Army’s SGT STAR – as well as a collection of brain-damaged case-studies the likes of which would fascinate even the pickiest Oliver Sacks fan. If you think you can’t be fooled by a cleverly-written script, the next time you’re desperate for tech help and Click to Chat! to a real-live expert, try getting a rise out of your deponent. Only if you can manage to get it off book and not spewing non sequiturs will you have assurances you have an intelligent counterpart. (This still might not be human, but at this level I think it prejudice to discriminate.)
The book overruns with ideas. For example, Christian offers a terrific alternative to the zero-sum forensics of debate/Model UN at pp. 179-180 that manages to promote constructive, persuasive argument and collaboration in a competitive context that all educators should be made to read. For another, there’s this brilliant passage at p. 237 that all Goodreaders ought to appreciate:
To put it another way: everything that has ever been written or will ever be written can be found in the Universal Library (including all languages, all typos, all lossless compressions of all 4D smellavision film versions, and this review). It has been said (by Martin Gardner, I believe, that it would take more matter than we know to actually be in existence to replicate the Library. On this basis alone, you’d think it safe to conclude that no computer could be devised to make appropriate selections… that each distinct work must be the result of an independently operating, organic intelligence. But computers are matter as we are matter, and we believe ourselves each to have a unique identity, a unique voice.
Brian Christian argues passionately that we can work at and improve our too-human, perfectly-flawed organic experience, and thereby exalt ourselves. We need never lose a *real* Turing test to a computer, however it may be run. Of course, computers, programmers' children, are wonderful mimics.
It's only a matter of time.
Those wishing to preview the themes and overall plot of the book in greater depth can read the author’s own 8-page online redaction. I did, and can assure you it only whetted my appetite to read the book itself. Nor did this book disappoint.
For those too time-strapped or lazy to check out the redacted version, here’s a thumbnail sketch. If you only had 5 minutes of text-only IM’ing to convince somebody you were YOU, could you do it? How would a stranger figure out that yours were not the automated responses of a preprogrammed ‘bot, nor that of a phisher, nor even that of David Bowie, but YOU. (Umm, uunless, of course, you *are* David Bowie, in which case, may I suggest you use this book’s premise for your next concept album?) Could you convince them? And how would you do it? How do we recognize each other as distinct individuals? How does one measure the soul? Who are we really?
That’s the challenge that Christian sets for himself, and if this strikes you as a bit arbitrary as personal challenges go, then you need a bit of background on Alan Turing’s test and its annual incarnation as the Loebner Prize.
Alan Turing was a math genius who cracked the Nazi’s Enigma code and helped invent the modern-day computer among other things. (This bit of history is fascinating and well worth your time, as you get to meet not only Turing, but guys like Claude Shannon, Johnny Von Neumann, John Nash, and others. Fiction readers can find them in Neal Stephenson’s Cryptonomicon and the film “A Beautiful Mind;” casual nonfiction readers will find them in books by William Poundstone, James Gleick, and Richard Rhodes, and all of these I highly recommend.
Turing was fascinated by the idea of artificial intelligence and thinking machines and proposed that the only way to tell the difference between a sentient and nonsentient being was through a basic conservatory audition. Only, instead of some aspiring floutist piping their way through Fauré’s Morceau des Concours in hopes of an appointment, it was an invisible conversant. If the auditors believed their interlocutor to be human, then voila! Whatever turned out to be behind the screen would have to be considered capable of independent thought.
Seem crazy? Yeah, crazy like a fox, and it was only a matter of time before the Turing test got institutionalized in the form of the Loebner Prize competition, that 5-minute IM thing to which Christian wangled himself an invitation (thus the challenge at the core of this book). However, Christian isn’t content to “just be himself” in the hope and expectation that so doing will prove sufficient to win over the Loebner’s panel of PhD’d judges (psychologists, linguists, computer programmers, etc.), nor does Christian want merely to outcompete the various chatbots entered alongside him and his fellow human controls/confederates. No, Christian is out to prove to all and sundry that not only is he capable of demonstrating his humanity anonymously to strangers, he can do so better than any other human in the competition. Much of the book deals with his efforts as a kind of Rocky Balboa IV, training himself to challenge the Ivan Dragos of chatbots and fellow panelists to a human-off.
Think it’s easy, do you? Try sitting down next to someone at random on the Acela and striking up a conversation without mentioning the weather or running some other personally pre-programmed bit of dialogue: “Hi, how are you, my name’s ___, where are you headed, etc.” Unless you’re a transactional analyst, you’d be surprised how much of our lives come to us pre-scripted in your typical five-minute span.
The issues touched on here run from the sublimity of learning to live life to its fullest (by avoiding rote, repetitive experiences) to the ridiculous informational density of language (he shares his results playing the “Shannon Game” at p. 227 -– I tried it and can tell you that the phrases given are bizarre, and not always grammatical; anyway, it’s worth playing once). Along the way, he offers a rogues gallery of chatbots -- from the ever-popular therapist Eliza and her patient foil Manny all the way to present-day AIs like the U.S. Army’s SGT STAR – as well as a collection of brain-damaged case-studies the likes of which would fascinate even the pickiest Oliver Sacks fan. If you think you can’t be fooled by a cleverly-written script, the next time you’re desperate for tech help and Click to Chat! to a real-live expert, try getting a rise out of your deponent. Only if you can manage to get it off book and not spewing non sequiturs will you have assurances you have an intelligent counterpart. (This still might not be human, but at this level I think it prejudice to discriminate.)
The book overruns with ideas. For example, Christian offers a terrific alternative to the zero-sum forensics of debate/Model UN at pp. 179-180 that manages to promote constructive, persuasive argument and collaboration in a competitive context that all educators should be made to read. For another, there’s this brilliant passage at p. 237 that all Goodreaders ought to appreciate:
People complain from time to time about folks who read the Cliffs-Notes to a book, or reviews or essays about a book, but don’t read the book itself. Hey, if the information density of Anna Karenina is low enough that a review 1 percent as long conveys 60 percent of the form and content ‘gist’ of the book, then it’s Tolstoy’s fault. His readers are human beings with only twenty-eight thousand days or so separating birth and death. If they want to read the lossy gloss and move on, who can blame them?The Most Human Human is really a wonderful book. You can take it from me, assuming that at this point in my exegesis I have indeed managed not only to bring you round, but have done so without you having recourse to wonder whether or not *I* and only I wrote this text -- without resort to plagiarism, paraphrase, or other outside assistance organic or otherwise. (See, e.g., my trilogy of essays on this conundrum, the first of which can be found here.)
Likewise for conceptual art: who needs to see a Duchamp toilet when you can hear about one so much faster and extract most of the experience from that? Conceptual art might be, for better or worse, (definable as) the art most susceptible to lossy compression.
To put it another way: everything that has ever been written or will ever be written can be found in the Universal Library (including all languages, all typos, all lossless compressions of all 4D smellavision film versions, and this review). It has been said (by Martin Gardner, I believe, that it would take more matter than we know to actually be in existence to replicate the Library. On this basis alone, you’d think it safe to conclude that no computer could be devised to make appropriate selections… that each distinct work must be the result of an independently operating, organic intelligence. But computers are matter as we are matter, and we believe ourselves each to have a unique identity, a unique voice.
Brian Christian argues passionately that we can work at and improve our too-human, perfectly-flawed organic experience, and thereby exalt ourselves. We need never lose a *real* Turing test to a computer, however it may be run. Of course, computers, programmers' children, are wonderful mimics.
It's only a matter of time.
November 23, 2022
Once in a while, we chanced upon a book that altered the way we look at things. The everyday things.
Learning from machines to become more human. Doesn’t it sound odd?
Brain Christian rightly pointed out that computers are a tool designed without a specified purpose initially, and tons of applications were designed after computers (tools) were constructed. Now, we can see the power of computing applied to almost every mundane activity.
With computers seemingly invading all territories of our daily activities, taking over human jobs, and with our ever-increasing reliance on computers to manage our schedule and tasks, inevitably, we question just how much has machines become human.
To answer that question, Brian Christian goes on to explore what makes a human.... human?
The differentiating factor is our unique ability to "communicate" with each other.
Most languages are designed around a very structured system, yet we adopt an unstructured way to deliver our ideas, thoughts, and feelings every day. Our tendency to add in conversational fillers (uhh, umm), having unique slang, the ability to have our point of view, having a 'stateful' identity, and adding in emotions, all are contributing factors that add depth and flow into a conversation that makes us human. Something computers just aren't quite there yet.
By creating new thoughts and ideas, and selectively choosing words, every conversation in itself is poetry. Conversations with machines are repetition and a mimic of ideas that are after all... inputs.
Without those 'human' elements, we will never get the form of satisfaction and fulfillment in a conversation with a computer.
The duels of arguments and counter-arguments.
The gamble of approvals and rejections.
The power of anger and encouragement.
The emotional energy that bounces off two people.
Conversation with another human being is very much rewarding. Something that computers will never take away from us. And I'll start to treasure the one thing that makes me uniquely human.
Learning from machines to become more human. Doesn’t it sound odd?
Brain Christian rightly pointed out that computers are a tool designed without a specified purpose initially, and tons of applications were designed after computers (tools) were constructed. Now, we can see the power of computing applied to almost every mundane activity.
With computers seemingly invading all territories of our daily activities, taking over human jobs, and with our ever-increasing reliance on computers to manage our schedule and tasks, inevitably, we question just how much has machines become human.
To answer that question, Brian Christian goes on to explore what makes a human.... human?
The differentiating factor is our unique ability to "communicate" with each other.
Most languages are designed around a very structured system, yet we adopt an unstructured way to deliver our ideas, thoughts, and feelings every day. Our tendency to add in conversational fillers (uhh, umm), having unique slang, the ability to have our point of view, having a 'stateful' identity, and adding in emotions, all are contributing factors that add depth and flow into a conversation that makes us human. Something computers just aren't quite there yet.
By creating new thoughts and ideas, and selectively choosing words, every conversation in itself is poetry. Conversations with machines are repetition and a mimic of ideas that are after all... inputs.
Without those 'human' elements, we will never get the form of satisfaction and fulfillment in a conversation with a computer.
The duels of arguments and counter-arguments.
The gamble of approvals and rejections.
The power of anger and encouragement.
The emotional energy that bounces off two people.
Conversation with another human being is very much rewarding. Something that computers will never take away from us. And I'll start to treasure the one thing that makes me uniquely human.
This entire review has been hidden because of spoilers.
March 10, 2018
Fantastic. Very readable and surprisingly interesting given how much of the referenced material I've already read.
April 27, 2023
วิเศษมาก! เป็นหนึ่งในหนังสือที่วิเศษที่สุดที่ได้อ่านในรอบหลายปีมานี้
Brian Christian ผู้เขียนหนังสือเล่มนี้คือผู้ชนะรางวัล “มนุษย์ที่เป็นมนุษย์ที่สุด” ในการทดสอบทัวริงเพื่อชิงรางวัลโลบเนอร์ในปี 2009 แต่นี่ไม่ใช่แค่หนังสือที่บอกเล่าประสบการณ์ในการแข่งขันกับโปรแกรมคอมพิวเตอร์หรือป��ญญาประดิษฐ์ หรือบอกกลเม็ดเคล็ดลับที่ทำให้เขาชนะในการแข่งขันได้ แต่ทั้งหมดที่หนังสือเล่มนี้เป็น อาจเรียกได้ว่ามันคือสรวงสวรรค์สำหรับคนที่หลงใหลในปรัชญา ภาษาศาสตร์ วรรณกรรม บทกวี cognitive science computer scine และอีกหลายศาสตร์ที่เคลื่อนผ่านคำถามว่าอะไรคือ “ความเป็นมนุษย์” และเป็นอุดมคติของวิชาสายมนุษยศาสตร์ที่พาเราออกไปไกลกว่ากรอบการอธิบายผ่าน “ภาษา” และ “วัฒนธรรม” มัน��ป็นทุกอย่างตั้งแต่ตำราสอนวิธีจีบสาวและการออกเดท บทวิเคราะห์กวีนิพนธ์ ศิลปะ และวรรณกรรม ที่เขียนโดยโปรแกรมเมอร์ที่เป็นกวี และกวีที่เป็นโปรแกรมเมอร์
ไม่เกินเลยนักถ้าจะบอกว่านี่คือ “คู่มือการเป็นมนุษย์” (ไม่ใช่ในแง่ศีลธรรม) ที่ละเอียด ลึกซึ้ง และรอบด้านมากที่สุดเล่มหนึ่ง ถ้าปัญญาประดิษฐ์สามารถเรียนรู้ได้ถึงขั้นที่ว่าความเป็นมนุษย์หมายความว่ายังไง หนังสือเล่มนี้ก็อาจเปรียบได้กับคู่มือที่จะ “สอน” ปัญญาประดิษฐ์ว่าความเป็นมนุษย์คืออะไร
ในฐานะคนที่พอรู้กระแสความคิดแบบ posthuman อยู่บ้าง และเ��็นการวิพากษ์วิธีคิดแบบ “มนุษยนิยม” มาพอสมควร เราก็คิดว่าหากวิธีคิดแบบมนุษยนิยมจะกลับมาทวงบัลลังก์อีกครั้ง มันก็น่าจะต้องนำเสนอตัวเองในแบบที่หนังสือเล่มนี้สาธิตให้เห็นตลอดทั้งเล่มนี่แหละ หาก “มนุษยนิยม” นิยามความเป็นมนุษย์แบบ exclusivism และ “หลังมนุษยนิยม” นิยามความเป็นมนุษย์แบบ inclusivism เราว่า “ความเป็นมนุษย์” ที่หนังสือเล่มนี้พยายามธิบาย คือสิ่งที่อยู่ตรงกลางระหว่างความบ้าคลั่งของคำนิยามสองสายนั้น
แปลกดีที่อ่านหนังสือเล่มนี้แล้วทำให้เรารู้สึกว่าจริงๆ แล้ว “มนุษย์” กับเครื่องจักร/หุ่นยนต์/ปัญญาประดิษฐ์ ไม่ได้แยกขาดจากกันหรือเป็นศัตรูกันขนาดนั้น เรากลับรู้สึกว่าทั้งสองเป็นเหมือน “ร่าง” กับ “เงา” ที่จะคู่ขนานกันไปเรื่อยๆ
เราอาจรู้สึกปนกันทั้งหวาดหวั่น หัวเราะ เอ็นดู ฯลฯ ที่เห็นเครื่องจักร/หุ่นยนต์/ปัญญาประดิษฐ์กลายเป็น “มนุษย์” มากขึ้น เหมือนได้เห็นตัวเองในเวอร์ชั่นที่ยังด้อยกว่า แต่เรากลับรู้น้อยเหลือเกินว่าเราจะเป็นเราในเวอร์ชั่นที่ดีกว่าได้ยังไง หรือนั่นอาจเป็นเพราะว่ามนุษย์ไม่มี “ต้นแบบ” ให้เลียนแบบ (เหมือนที่เครื่องจักรเลียนแบบความเป็นมนุษย์) และการไม่มีต้นแบบของเราก็ช่วยมอบความเป็นไปได้ที่ไม่รู้จบ
Brian Christian ผู้เขียนหนังสือเล่มนี้คือผู้ชนะรางวัล “มนุษย์ที่เป็นมนุษย์ที่สุด” ในการทดสอบทัวริงเพื่อชิงรางวัลโลบเนอร์ในปี 2009 แต่นี่ไม่ใช่แค่หนังสือที่บอกเล่าประสบการณ์ในการแข่งขันกับโปรแกรมคอมพิวเตอร์หรือป��ญญาประดิษฐ์ หรือบอกกลเม็ดเคล็ดลับที่ทำให้เขาชนะในการแข่งขันได้ แต่ทั้งหมดที่หนังสือเล่มนี้เป็น อาจเรียกได้ว่ามันคือสรวงสวรรค์สำหรับคนที่หลงใหลในปรัชญา ภาษาศาสตร์ วรรณกรรม บทกวี cognitive science computer scine และอีกหลายศาสตร์ที่เคลื่อนผ่านคำถามว่าอะไรคือ “ความเป็นมนุษย์” และเป็นอุดมคติของวิชาสายมนุษยศาสตร์ที่พาเราออกไปไกลกว่ากรอบการอธิบายผ่าน “ภาษา” และ “วัฒนธรรม” มัน��ป็นทุกอย่างตั้งแต่ตำราสอนวิธีจีบสาวและการออกเดท บทวิเคราะห์กวีนิพนธ์ ศิลปะ และวรรณกรรม ที่เขียนโดยโปรแกรมเมอร์ที่เป็นกวี และกวีที่เป็นโปรแกรมเมอร์
ไม่เกินเลยนักถ้าจะบอกว่านี่คือ “คู่มือการเป็นมนุษย์” (ไม่ใช่ในแง่ศีลธรรม) ที่ละเอียด ลึกซึ้ง และรอบด้านมากที่สุดเล่มหนึ่ง ถ้าปัญญาประดิษฐ์สามารถเรียนรู้ได้ถึงขั้นที่ว่าความเป็นมนุษย์หมายความว่ายังไง หนังสือเล่มนี้ก็อาจเปรียบได้กับคู่มือที่จะ “สอน” ปัญญาประดิษฐ์ว่าความเป็นมนุษย์คืออะไร
ในฐานะคนที่พอรู้กระแสความคิดแบบ posthuman อยู่บ้าง และเ��็นการวิพากษ์วิธีคิดแบบ “มนุษยนิยม” มาพอสมควร เราก็คิดว่าหากวิธีคิดแบบมนุษยนิยมจะกลับมาทวงบัลลังก์อีกครั้ง มันก็น่าจะต้องนำเสนอตัวเองในแบบที่หนังสือเล่มนี้สาธิตให้เห็นตลอดทั้งเล่มนี่แหละ หาก “มนุษยนิยม” นิยามความเป็นมนุษย์แบบ exclusivism และ “หลังมนุษยนิยม” นิยามความเป็นมนุษย์แบบ inclusivism เราว่า “ความเป็นมนุษย์” ที่หนังสือเล่มนี้พยายามธิบาย คือสิ่งที่อยู่ตรงกลางระหว่างความบ้าคลั่งของคำนิยามสองสายนั้น
แปลกดีที่อ่านหนังสือเล่มนี้แล้วทำให้เรารู้สึกว่าจริงๆ แล้ว “มนุษย์” กับเครื่องจักร/หุ่นยนต์/ปัญญาประดิษฐ์ ไม่ได้แยกขาดจากกันหรือเป็นศัตรูกันขนาดนั้น เรากลับรู้สึกว่าทั้งสองเป็นเหมือน “ร่าง” กับ “เงา” ที่จะคู่ขนานกันไปเรื่อยๆ
เราอาจรู้สึกปนกันทั้งหวาดหวั่น หัวเราะ เอ็นดู ฯลฯ ที่เห็นเครื่องจักร/หุ่นยนต์/ปัญญาประดิษฐ์กลายเป็น “มนุษย์” มากขึ้น เหมือนได้เห็นตัวเองในเวอร์ชั่นที่ยังด้อยกว่า แต่เรากลับรู้น้อยเหลือเกินว่าเราจะเป็นเราในเวอร์ชั่นที่ดีกว่าได้ยังไง หรือนั่นอาจเป็นเพราะว่ามนุษย์ไม่มี “ต้นแบบ” ให้เลียนแบบ (เหมือนที่เครื่องจักรเลียนแบบความเป็นมนุษย์) และการไม่มีต้นแบบของเราก็ช่วยมอบความเป็นไปได้ที่ไม่รู้จบ
April 20, 2016
This was an interesting read. Not for the reasons I was expecting when I started it, but interesting nonetheless. Christian starts out with his raison d'ecrire (hah..try Shannon Gaming that one!) as the Loebner prize 2009 Turning test competition, but quickly veers off into a discussion of, well lots of stuff. As other reviewers have noted, he does not always follow a tangential line of thought and it is not always clear how all of his details are relevant. Nor does he really describe the Loebner prize.
Instead, he gives us a history of AI and computer science and a philosophical discussion about what it means to be a human (from the "I think therefore I am" POV all the way to "we don't parse our words when we speak aloud"). The discussion was interesting (and yeah, I get that he needed to write a full book and there is lots to meat up just the competition), but I kinda wanted to know more about the actual Turing test. I would have liked transcripts of THE ACTUAL COMPETITION for example.
Despite not always relating, Christian is pretty easy to read and I learned some from the book, so it gets 4 stars. One of my chuckle aloud moments was when reading his footnote about how he always enjoys the beginnings of books more than the end. At some point, the completions of the project (finish the book) becomes more important than the book itself and he blames this (in part) on the book software he uses to review/take notes on what he reads. Hah! I can totally relate to getting to a point in a book where I know I am gonna finish it today or tonight and then HAVE TO WRITE A REVIEW. And yeah, I am eager to just get done with it already so I can start a new one. One that is bright and sparkly and doesn't have a pending review in the next few hours.
I have below my favorite moments/comments/soundbytes:
"Having a sense of a person--their disposition, character, 'way of being in the world'--and knowing about them--where they grew up, how many siblings they have, what they majored in, where they work--are two rather different things."
"Our very essence is a kind of mongrelism. It strikes me that some of the bets and most human emotions come from this lichen state of computer/creature interface, the admixture, the estuary of desire and reason in a system aware enough to apprehend its own limits, and to push at them: curiosity, intrigue, enlightenment, wonder, awe."
"I suppose when you get down to it, everything is always once in a lifetime. We might as well act like it."
"It's what we want, chatting with old friends, when our familiar opening book....is not so much a conversation per se as a means for arriving at one-gives pleasantly way to the expectedly unexpected, awaitedly idiosyncratic veers; it's what anyone wants from any conversation, and what artist want from their art; a way to breeze past formalities and received gestures, out of book and into the real thing."
"computers become in effect the first tools to precede their tasks: their fundamental difference from staplers and hole-punchers and pocket watches. You build the computer first, and then figure out what you want to do.....seems to chip away at the existentialist idea of humans' unique purchase on the idea of existence before essence."
Instead, he gives us a history of AI and computer science and a philosophical discussion about what it means to be a human (from the "I think therefore I am" POV all the way to "we don't parse our words when we speak aloud"). The discussion was interesting (and yeah, I get that he needed to write a full book and there is lots to meat up just the competition), but I kinda wanted to know more about the actual Turing test. I would have liked transcripts of THE ACTUAL COMPETITION for example.
Despite not always relating, Christian is pretty easy to read and I learned some from the book, so it gets 4 stars. One of my chuckle aloud moments was when reading his footnote about how he always enjoys the beginnings of books more than the end. At some point, the completions of the project (finish the book) becomes more important than the book itself and he blames this (in part) on the book software he uses to review/take notes on what he reads. Hah! I can totally relate to getting to a point in a book where I know I am gonna finish it today or tonight and then HAVE TO WRITE A REVIEW. And yeah, I am eager to just get done with it already so I can start a new one. One that is bright and sparkly and doesn't have a pending review in the next few hours.
I have below my favorite moments/comments/soundbytes:
"Having a sense of a person--their disposition, character, 'way of being in the world'--and knowing about them--where they grew up, how many siblings they have, what they majored in, where they work--are two rather different things."
"Our very essence is a kind of mongrelism. It strikes me that some of the bets and most human emotions come from this lichen state of computer/creature interface, the admixture, the estuary of desire and reason in a system aware enough to apprehend its own limits, and to push at them: curiosity, intrigue, enlightenment, wonder, awe."
"I suppose when you get down to it, everything is always once in a lifetime. We might as well act like it."
"It's what we want, chatting with old friends, when our familiar opening book....is not so much a conversation per se as a means for arriving at one-gives pleasantly way to the expectedly unexpected, awaitedly idiosyncratic veers; it's what anyone wants from any conversation, and what artist want from their art; a way to breeze past formalities and received gestures, out of book and into the real thing."
"computers become in effect the first tools to precede their tasks: their fundamental difference from staplers and hole-punchers and pocket watches. You build the computer first, and then figure out what you want to do.....seems to chip away at the existentialist idea of humans' unique purchase on the idea of existence before essence."
July 11, 2011
A book exploring the wild frontiers of chat-bots is appealing enough; I never expected to discover in its pages such an eye-opening inquest into human imagination, thought, conversation, love and deception. Who would have guessed that the best way to understand humanity was to study its imitators?
October 1, 2014
Avete mai sentito parlare del premio Loebner? No? Allora questo è il libro che fa per voi! Si tratta di una competizione organizzata dall’imprenditore Hugh Loebner e aperta ai programmatori di Artificial Intelligence (d’ora in poi AI): i partecipanti sono programmi di AI in grado di sostenere una conversazione con esseri umani, meglio noti come chatterbot o chatbot, e devono superare una particolare versione del cosiddetto test di Turing. Il matematico britannico Alan M. Turing, una delle menti più brillanti del XX secolo, è considerato l’ideatore dei moderni concetti di computer e di AI. In un articolo del 1950, intitolato “Computing Machinery and Intelligence”, propose un test per valutare il grado di assimilazione del pensiero umano da parte di un’AI: un individuo - il giudice - sostiene una conversazione di pochi minuti a turno con due interlocutori a lui nascosti, un essere umano e un’AI; in base al dialogo il giudice deve decidere chi è l’umano. Turing sosteneva che prima o poi le macchine avrebbero raggiunto un grado di assimilazione tale da venir scambiate per esseri umani, provando così la loro capacità di pensare. Come è facile comprendere, si tratta di un tema scientifico, tecnologico e filosofico di estremo fascino.
Nella versione di Loebner, che organizzò la prima gara nel 1990, le conversazioni durano cinque minuti e il premio di 100,000 dollari viene assegnato ai programmatori dell’AI che più si è avvicinata alla soglia del 30% di giudici “ingannati” (che scambiano l’AI per un essere umano): Turing riteneva tale soglia un obiettivo raggiungibile nel 2000. Il giovane Brian Christian, poeta e laureato in informatica e filosofia (eh sì, negli USA tutto ciò è possibile!), ma anche giornalista e divulgatore scientifico, ha partecipato in qualità di “associato” (confederate) ossia di controparte umana dell’AI, nella competizione del 2009: il suo obiettivo era però farsi riconoscere come l’umano “più umano” - da cui il titolo originale del libro: “The Most Human Human” - e ricevere così il relativo premio, destinato all’associato che acquista il maggior numero di attestati di umanità da parte del giudici.
Lascio scoprire a chi vorrà leggere il libro se Christian è riuscito nell’intento. Quanto al libro, bisogna dire che è davvero un lavoro divulgativo molto interessante ed estremamente stimolante: esso ruota attorno a due fuochi, le potenzialità dell’AI nel simulare il comportamento umano e l’identificazione di ciò che qualifica come umano quello stesso comportamento. Partendo dal premio e dalle sue complesse modalità di svolgimento (criticate da filosofi e scienziati), Christian intraprende un viaggio alla scoperta di quelle caratteristiche che rendono l’essere umano così particolare e che dovrebbero essere imitate da un’AI che volesse farsi passare per uomo o donna. Nel contempo affronta tematiche e problemi di matematica, filosofia, neuroscienze, filosofia del linguaggio e scienza dell’informazione: la coscienza e il sé, la mente e il cervello, le differenze tra i due emisferi, gli agenti razionali e la teoria dei giochi, le macchine di Turing, le tecniche di problem solving, la teoria della calcolabilità, la teoria della complessità, l’entropia informazionale. Christian non esista a connettere le grandi domande (che cos’è la coscienza? gli animali hanno un’anima? potremmo mai innamorarci di un computer?) a questioni più spicciole (che cos’è lo speed date? perché gli operatori dei call center hanno una spiacevole tendenza a dare risposte impersonali tipiche di un automa?). Si diverte molto a spiegarci come funziona un software per giocare a scacchi (cap. 4, una delle parti migliori del libro), che cosa erano le famose “calcolatrici” (una delle quali divenne moglie di Claude Shannon), qual è l’importanza delle pause e delle espressioni “Uh…”, “Ehm…” e “Beh…” in una conversazione, come si struttura la dialettica avvocato-testimone etc. etc. Se si nutre una forte curiosità per interrogativi molto eterogenei nel campo dell’AI, il libro di Christian può rivelarsi davvero attraente.
Il tono generale è piuttosto leggero e colloquiale, zeppo di aneddoti (telefonate e viaggi Interrail, marcatamente autoironici), giochi di parole, citazioni e riferimenti culturali anglosassoni (al cinema, alle serie tv, alla musica e alla letteratura contemporanee) talvolta non facilmente identificabili per il lettore italiano: tutto sommato è uno stile che rende i temi trattati più abbordabili, abbastanza efficace dal punto di vista divulgativo, anche se talvolta un po’ eccessivamente semplificatorio. Ad esempio nelle parti dedicate ai temi più filosofici, Christian tende a essere sbrigativo e mescola con disinvoltura Platone e Aristotele, lo stoicismo e il cristianesimo. Non tutte le citazioni sono perspicue e talvolta si ha l’impressione che Christian si lasci trasportare in voli pindarici non sempre giustificati. La veste grafica del libro (ripresa nella traduzione italiana) non è delle migliori e le citazioni non sono sufficientemente isolate dal testo. Nell’ultimo capitolo (il nono) Christian mette davvero troppa carne al fuoco - etica, linguistica, teoria dell’informazione, tecniche pubblicitarie -, finendo per perdersi senza approfondire niente: in effetti questo è un capitolo edito autonomamente e si amalgama male con il resto del libro. Si sente anche la mancanza di un indice analitico.
Le pecche imputabili all’autore sono comunque scusabili se confrontate con quelle dell’edizione italiana, che, mi duole dirlo, è piuttosto carente. Ho acquistato il libro come supplemento del mensile Le Scienze qualche tempo fa: sono un affezionato lettore di questa rivista e acquisto spessissimo i libri che mensilmente l’accompagnano e che contribuiscono a far uscire il nostro paese dall’analfabetismo matematico-scientifico che lo caratterizza. Nel caso in questione il lavoro di revisione non ha dato i frutti sperati, nonostante la traduzione sia stata realizzata da ben tre persone. A parte i numerosissimi refusi, le costruzioni involute, le improprietà sintattiche, vi si trovano vere e proprie cantonate, difficilmente accettabili: per fare qualche esempio, a pag. 69 “ever-feasting nobility” (“nobiltà impegnata in continui banchetti”) è tradotto come “nobilità lussureggiante” (?), a pag. 134 “teleological” è stato reso con “teologico” (“la vita più teologica, orientata verso un fine”). Mi è capitato di dover ricercare alcuni passaggi nell’edizione originale per capirne il senso. Per qualche ignoto motivo è stata eliminata la bibliografia, rendendo quindi impossibile risalire alle citazioni del testo, e le note ai singoli capitoli sono state riunite a fine libro. Suggerisco a chi ne ha la possibilità di leggerlo in inglese.
Consigliato a chi apprezza il piacere di una buona conversazione.
Sconsigliato a Sarah Connor.
Nella versione di Loebner, che organizzò la prima gara nel 1990, le conversazioni durano cinque minuti e il premio di 100,000 dollari viene assegnato ai programmatori dell’AI che più si è avvicinata alla soglia del 30% di giudici “ingannati” (che scambiano l’AI per un essere umano): Turing riteneva tale soglia un obiettivo raggiungibile nel 2000. Il giovane Brian Christian, poeta e laureato in informatica e filosofia (eh sì, negli USA tutto ciò è possibile!), ma anche giornalista e divulgatore scientifico, ha partecipato in qualità di “associato” (confederate) ossia di controparte umana dell’AI, nella competizione del 2009: il suo obiettivo era però farsi riconoscere come l’umano “più umano” - da cui il titolo originale del libro: “The Most Human Human” - e ricevere così il relativo premio, destinato all’associato che acquista il maggior numero di attestati di umanità da parte del giudici.
Lascio scoprire a chi vorrà leggere il libro se Christian è riuscito nell’intento. Quanto al libro, bisogna dire che è davvero un lavoro divulgativo molto interessante ed estremamente stimolante: esso ruota attorno a due fuochi, le potenzialità dell’AI nel simulare il comportamento umano e l’identificazione di ciò che qualifica come umano quello stesso comportamento. Partendo dal premio e dalle sue complesse modalità di svolgimento (criticate da filosofi e scienziati), Christian intraprende un viaggio alla scoperta di quelle caratteristiche che rendono l’essere umano così particolare e che dovrebbero essere imitate da un’AI che volesse farsi passare per uomo o donna. Nel contempo affronta tematiche e problemi di matematica, filosofia, neuroscienze, filosofia del linguaggio e scienza dell’informazione: la coscienza e il sé, la mente e il cervello, le differenze tra i due emisferi, gli agenti razionali e la teoria dei giochi, le macchine di Turing, le tecniche di problem solving, la teoria della calcolabilità, la teoria della complessità, l’entropia informazionale. Christian non esista a connettere le grandi domande (che cos’è la coscienza? gli animali hanno un’anima? potremmo mai innamorarci di un computer?) a questioni più spicciole (che cos’è lo speed date? perché gli operatori dei call center hanno una spiacevole tendenza a dare risposte impersonali tipiche di un automa?). Si diverte molto a spiegarci come funziona un software per giocare a scacchi (cap. 4, una delle parti migliori del libro), che cosa erano le famose “calcolatrici” (una delle quali divenne moglie di Claude Shannon), qual è l’importanza delle pause e delle espressioni “Uh…”, “Ehm…” e “Beh…” in una conversazione, come si struttura la dialettica avvocato-testimone etc. etc. Se si nutre una forte curiosità per interrogativi molto eterogenei nel campo dell’AI, il libro di Christian può rivelarsi davvero attraente.
Il tono generale è piuttosto leggero e colloquiale, zeppo di aneddoti (telefonate e viaggi Interrail, marcatamente autoironici), giochi di parole, citazioni e riferimenti culturali anglosassoni (al cinema, alle serie tv, alla musica e alla letteratura contemporanee) talvolta non facilmente identificabili per il lettore italiano: tutto sommato è uno stile che rende i temi trattati più abbordabili, abbastanza efficace dal punto di vista divulgativo, anche se talvolta un po’ eccessivamente semplificatorio. Ad esempio nelle parti dedicate ai temi più filosofici, Christian tende a essere sbrigativo e mescola con disinvoltura Platone e Aristotele, lo stoicismo e il cristianesimo. Non tutte le citazioni sono perspicue e talvolta si ha l’impressione che Christian si lasci trasportare in voli pindarici non sempre giustificati. La veste grafica del libro (ripresa nella traduzione italiana) non è delle migliori e le citazioni non sono sufficientemente isolate dal testo. Nell’ultimo capitolo (il nono) Christian mette davvero troppa carne al fuoco - etica, linguistica, teoria dell’informazione, tecniche pubblicitarie -, finendo per perdersi senza approfondire niente: in effetti questo è un capitolo edito autonomamente e si amalgama male con il resto del libro. Si sente anche la mancanza di un indice analitico.
Le pecche imputabili all’autore sono comunque scusabili se confrontate con quelle dell’edizione italiana, che, mi duole dirlo, è piuttosto carente. Ho acquistato il libro come supplemento del mensile Le Scienze qualche tempo fa: sono un affezionato lettore di questa rivista e acquisto spessissimo i libri che mensilmente l’accompagnano e che contribuiscono a far uscire il nostro paese dall’analfabetismo matematico-scientifico che lo caratterizza. Nel caso in questione il lavoro di revisione non ha dato i frutti sperati, nonostante la traduzione sia stata realizzata da ben tre persone. A parte i numerosissimi refusi, le costruzioni involute, le improprietà sintattiche, vi si trovano vere e proprie cantonate, difficilmente accettabili: per fare qualche esempio, a pag. 69 “ever-feasting nobility” (“nobiltà impegnata in continui banchetti”) è tradotto come “nobilità lussureggiante” (?), a pag. 134 “teleological” è stato reso con “teologico” (“la vita più teologica, orientata verso un fine”). Mi è capitato di dover ricercare alcuni passaggi nell’edizione originale per capirne il senso. Per qualche ignoto motivo è stata eliminata la bibliografia, rendendo quindi impossibile risalire alle citazioni del testo, e le note ai singoli capitoli sono state riunite a fine libro. Suggerisco a chi ne ha la possibilità di leggerlo in inglese.
Consigliato a chi apprezza il piacere di una buona conversazione.
Sconsigliato a Sarah Connor.
January 25, 2012
I didn't know what to expect from this book, but it surprised and delighted me with its thoughtful but not stodgy exploration of what it means to be human. The author entered the annual staging of the Turing Test--not as an author of a chatbot, but as a human. The Turing Test is where judges blindly IM with chatbots and humans and try to tell them apart; if a chatbot is reliably mistaken for human, the creator of the test proposed, then it could be said to be artificially intelligent. The book explores the nature of human intelligence and of artificial attempts at intelligence, from the significance of chess to competitive chatbots to neuroscience and philosophy.
Think of the Turing Test. If a computer program wins, that means that a human lost: judges have two conversations at a time and must decide which is the human and which is the computer. If a program is declared to be human, that means the human was declared to be a computer. The human lacked ... humanity? The advice to human participants is to "just be yourself", but the author can't do that. He's an over-achiever, he wants to study for and excel at this test of his humanity.
The topics weren't so enchanting as the approach; the author dazzled me with insight and connection on every page. I know a little about Artificial Intelligence, I thought: I took a class at uni, I pushed machine learning on O'Reilly Radar before everyone was doing it, I've written a chatbot, and I've read Douglas Hofstadter. But, not just once but repeatedly and consistently, Christian surprises me with a new way of looking at a familiar topic.
For example:
I'm familiar with this line of thinking, but this is the first time I'd seen it presented in a way that let me connect it directly to the "God of The Gaps" argument. In that line of thought, God is presumed to be behind everything that science can't explain, which leaves God to be "what science hasn't explained yet", presumably on a monotonically decreasing course to irrelevance. Similarly, the Thing That Makes Humans Special (what Harvard psychologist Daniel Gilbert calls "The Sentence": "The human being is the only animal that can ___") is constantly being shot down and replaced with something new:
I really liked the section on computer chess. I had forgotten how seriously chess was held up as an icon of man's superiority over the machine, even though now that attitude seems a quaint relic of a bygone day. Particularly eye-opening was how IBM refused a rematch: as soon as they'd won (through a Kasparov error that was the equivalent of poker's throwing away the pair and keeping the mismatched cards), they dismantled Deep Blue, ended the research program, and moved on. So they would play rematches, improving the computer each time, until man and computer were close enough that the computer could win one series ... at which point they denied the human the chance to improve and regain superiority. Didn't seem right.
He connects the strange conversational experiment of the Turing Test to speed dating (having to create a quick instant rapport with someone), Chat Roulette (one popular chatbot may work in a similar fashion, connecting humans to each other), philosophical insistence on "authenticity", coherence of identity (bots that learn responses from lots of people have no consistent self: will say they're a single woman when replying to one question, but happily married man as a reply to a different question), the Hardy Boy (written by a team of authors, not a single authorial voice), and finally:
It's this charming pan across the range of life and experience, showing how the dilemma of humanity runs through each vignette, that makes this book a keeper. If it had kept up the delight and glory of the first third, I'd have given it five stars. To be fair, though, I did stay up to a ridiculously late/early hour to finish this, so flaws in the last two thirds may well be entirely in my mind. And that, I must say, seems fitting.
Think of the Turing Test. If a computer program wins, that means that a human lost: judges have two conversations at a time and must decide which is the human and which is the computer. If a program is declared to be human, that means the human was declared to be a computer. The human lacked ... humanity? The advice to human participants is to "just be yourself", but the author can't do that. He's an over-achiever, he wants to study for and excel at this test of his humanity.
The topics weren't so enchanting as the approach; the author dazzled me with insight and connection on every page. I know a little about Artificial Intelligence, I thought: I took a class at uni, I pushed machine learning on O'Reilly Radar before everyone was doing it, I've written a chatbot, and I've read Douglas Hofstadter. But, not just once but repeatedly and consistently, Christian surprises me with a new way of looking at a familiar topic.
For example:
“Sometimes it seems,” says Douglas Hofstadter, “as though each new step towards AI, rather than producing something which everyone agrees is real intelligence, merely reveals what real intelligence is not.” While at first this seems a consoling position—one that keeps our unique claim to thought intact—it does bear the uncomfortable appearance of a gradual retreat, the mental image being that of a medieval army withdrawing from the castle to the keep. But the retreat can’t continue indefinitely. Consider: if everything of which we regarded “thinking” to be a hallmark turns out not to involve it, then ... what is thinking? It would seem to reduce to either an epiphenomenon—a kind of “exhaust” thrown off by the brain—or, worse, an illusion.
I'm familiar with this line of thinking, but this is the first time I'd seen it presented in a way that let me connect it directly to the "God of The Gaps" argument. In that line of thought, God is presumed to be behind everything that science can't explain, which leaves God to be "what science hasn't explained yet", presumably on a monotonically decreasing course to irrelevance. Similarly, the Thing That Makes Humans Special (what Harvard psychologist Daniel Gilbert calls "The Sentence": "The human being is the only animal that can ___") is constantly being shot down and replaced with something new:
And here’s a crucial, related question: Is this retreat a good thing or a bad thing? For instance, does the fact that computers are so good at mathematics in some sense take away an arena of human activity, or does it free us from having to do a nonhuman activity, liberating us into a more human life? The latter view would seem to be the more appealing, but it starts to seem less so if we can imagine a point in the future where the number of “human activities” left to be “liberated” into has grown uncomfortably small. What then?
I really liked the section on computer chess. I had forgotten how seriously chess was held up as an icon of man's superiority over the machine, even though now that attitude seems a quaint relic of a bygone day. Particularly eye-opening was how IBM refused a rematch: as soon as they'd won (through a Kasparov error that was the equivalent of poker's throwing away the pair and keeping the mismatched cards), they dismantled Deep Blue, ended the research program, and moved on. So they would play rematches, improving the computer each time, until man and computer were close enough that the computer could win one series ... at which point they denied the human the chance to improve and regain superiority. Didn't seem right.
He connects the strange conversational experiment of the Turing Test to speed dating (having to create a quick instant rapport with someone), Chat Roulette (one popular chatbot may work in a similar fashion, connecting humans to each other), philosophical insistence on "authenticity", coherence of identity (bots that learn responses from lots of people have no consistent self: will say they're a single woman when replying to one question, but happily married man as a reply to a different question), the Hardy Boy (written by a team of authors, not a single authorial voice), and finally:
The New York Times reported in June 2010—in an article titled “The End of the Best Friend”—on the practice of deliberate intervention, on the part of well-meaning adults, to disrupt close nuclei of friends from forming in schools and summer camps.4 One sleepaway camp in New York State, they wrote, has hired “friendship coaches” whose job is to notice whether “two children seem to be too focused on each other, [and] ... put them on different sports teams [or] seat them at different ends of the dining table.” Affirms one school counselor in St. Louis, “I think it is kids’ preference to pair up and have that one best friend. As adults—teachers and counselors—we try to encourage them not to do that.” Chatroulette and Omegle users “next” each other when the conversation flags; these children are being nexted by force—when things are going too well.
It's this charming pan across the range of life and experience, showing how the dilemma of humanity runs through each vignette, that makes this book a keeper. If it had kept up the delight and glory of the first third, I'd have given it five stars. To be fair, though, I did stay up to a ridiculously late/early hour to finish this, so flaws in the last two thirds may well be entirely in my mind. And that, I must say, seems fitting.
March 16, 2017
A competition called the Turing test takes place each year. Judges at computer terminals interact with unseen correspondents. Each judge has two correspondents, one a human being and one a computer program, and the judge tries to tell which is which after a five minute online conversation with each. The program that receives the most votes and highest judge confidence score is named the Most Human Computer. This title is highly coveted by programmers. A side result of the voting, however, is that the human who receives the most votes and highest judge confidence score is named the Most Human Human.
It is from this side of the Turing test that author Brian Christian writes his book The Most Human Human. He sets out to participate in the test as a correspondent and to win the Most Human Human title. Along the way, he philosophizes about what it means to be human and how our interaction with computers is affecting that. He notes, “We once thought humans were unique for having a language with syntactical rules, but this isn’t so; we once thought humans were unique for using tools, but this isn’t so; we once thought humans were unique for being able to do mathematics, and now we can barely imagine being able to do what our calculators can.”
The author makes the point that cell phones, texting, and programs that finish our words for us are making us less creative. It is easier to use the word the phone suggests than to fight the phone and type the word we meant to use. He writes, “I was detachedly roaming the Internet, but there was nothing interesting happening in the news, nothing interesting happening on Facebook…I grew despondent, depressed – the world used to seem so interesting…But all of a sudden it dawned on me, as if the thought had just occurred to me, that much of what is interesting and amazing about this world did not happen in the past twenty-four hours. How had this fact slipped away from me? …. Somehow I think the Internet is making this very critical point lost on an entire demographic.”
Christian is an interesting guy. He has a dual bachelor’s degree in computer science and philosophy and a master of fine arts in poetry. He understands the scientific angle of the Turing test but also the human side of what it means for a human to challenge a computer. There is a wonderful scene during the Turing test when he spies on a fellow human correspondent’s chat with a judge and realizes they are chatting in shorthand about Canadian hockey teams, virtually assuring that the judge knows he is talking to a human. This causes Christian a moment of panic and despair when he fears that he will lose the Most Human Human title.
Christian’s views on how we interact with the world are refreshing. He says, “I think the reason novels are regarded to have so much more ‘information’ than films is that they outsource the scenic design and cinematography to the reader. … This, for me, is a powerful argument for the value and potency of literature specifically.” I felt somewhat lost toward the end of the book when it got a bit scientific, but the science was not too overwhelming, and I wouldn’t let that put you off as a potential reader. I enjoyed this book tremendously, and it really made me think.
It is from this side of the Turing test that author Brian Christian writes his book The Most Human Human. He sets out to participate in the test as a correspondent and to win the Most Human Human title. Along the way, he philosophizes about what it means to be human and how our interaction with computers is affecting that. He notes, “We once thought humans were unique for having a language with syntactical rules, but this isn’t so; we once thought humans were unique for using tools, but this isn’t so; we once thought humans were unique for being able to do mathematics, and now we can barely imagine being able to do what our calculators can.”
The author makes the point that cell phones, texting, and programs that finish our words for us are making us less creative. It is easier to use the word the phone suggests than to fight the phone and type the word we meant to use. He writes, “I was detachedly roaming the Internet, but there was nothing interesting happening in the news, nothing interesting happening on Facebook…I grew despondent, depressed – the world used to seem so interesting…But all of a sudden it dawned on me, as if the thought had just occurred to me, that much of what is interesting and amazing about this world did not happen in the past twenty-four hours. How had this fact slipped away from me? …. Somehow I think the Internet is making this very critical point lost on an entire demographic.”
Christian is an interesting guy. He has a dual bachelor’s degree in computer science and philosophy and a master of fine arts in poetry. He understands the scientific angle of the Turing test but also the human side of what it means for a human to challenge a computer. There is a wonderful scene during the Turing test when he spies on a fellow human correspondent’s chat with a judge and realizes they are chatting in shorthand about Canadian hockey teams, virtually assuring that the judge knows he is talking to a human. This causes Christian a moment of panic and despair when he fears that he will lose the Most Human Human title.
Christian’s views on how we interact with the world are refreshing. He says, “I think the reason novels are regarded to have so much more ‘information’ than films is that they outsource the scenic design and cinematography to the reader. … This, for me, is a powerful argument for the value and potency of literature specifically.” I felt somewhat lost toward the end of the book when it got a bit scientific, but the science was not too overwhelming, and I wouldn’t let that put you off as a potential reader. I enjoyed this book tremendously, and it really made me think.
December 4, 2013
I seem to have a thing going for books that describe how the author tries to achieve some weird self-set goal (reading the Britannica in order to win a game show, reading through the complete OED, becoming US memory champion). In this instance the goal is to win the 'most human human' award that is given out at an annual Turing test.
To bring everyone up to speed: the Turing test was meant as a measure how well artificial intelligences perform. Judges have to have a 1:1 chat with a computer and a human confederate without knowing which 'intelligence' is only based in silico and have to make a call of which participant was human and which was the machine. If a computer can fool ~30% of the judges into believing that they are the human participant the program is said to have passed the Turing test and for the Loebner Prize there's the 'Most Human Computer'-Award for the program hat fools most judges, even if not passing the threshold. But as the competition is a zero-sum game (for each time the computer fools a judge into believing that it's human a human must be flagged as non-human) there's also the award for the human confederate that can convince the most judges that they are in fact human.
So that's what Christian is trying to win. And instead of just acting like a regular human being he sets out to prepare for the test by reading up on AI, computational linguistics and information theory to find the weak spots of current chatbots. Which is fun to read and one quickly notices that he has a professional background in computer science, philosophy and poetry. I really enjoyed reading about the different strategies applied to fake being human, like turning the chat into a shouting match, as insults are no longer dependent on a larger context but from a computer science perspective are more of a Markov chain, where your retort is only dependent on the last observed state.
Another fun topic was the application of the Shannon Entropy to text input on mobile devices via T9 and the different auto-correction algorithms around. Christian argues that these methods are shaping how you write, as they perform best on the vocabulary they were designed on and you thus unconsciously start to adapt to that vocabulary. Anecdotally this is just what I observe while typing away on my phone: Instead of the software adapting to my needs and figuring out what I wanted to say, I start to adapt to the idiosyncrasies of the auto-completion feature and try to predict what words the software can make most sense of, in order to avoid the software turning my intended writing into gibberish (especially for those üseless umlauts while typing in German. And I still have fond memories of T9 turning 'Sure, I'll come over and bring beer' into 'Sure, I'll come over and bring AIDS').
To borrow from Philipp's closings: Recommended for anyone interested in AI and computational linguistics (probably less if you're an expert in one of those fields).
To bring everyone up to speed: the Turing test was meant as a measure how well artificial intelligences perform. Judges have to have a 1:1 chat with a computer and a human confederate without knowing which 'intelligence' is only based in silico and have to make a call of which participant was human and which was the machine. If a computer can fool ~30% of the judges into believing that they are the human participant the program is said to have passed the Turing test and for the Loebner Prize there's the 'Most Human Computer'-Award for the program hat fools most judges, even if not passing the threshold. But as the competition is a zero-sum game (for each time the computer fools a judge into believing that it's human a human must be flagged as non-human) there's also the award for the human confederate that can convince the most judges that they are in fact human.
So that's what Christian is trying to win. And instead of just acting like a regular human being he sets out to prepare for the test by reading up on AI, computational linguistics and information theory to find the weak spots of current chatbots. Which is fun to read and one quickly notices that he has a professional background in computer science, philosophy and poetry. I really enjoyed reading about the different strategies applied to fake being human, like turning the chat into a shouting match, as insults are no longer dependent on a larger context but from a computer science perspective are more of a Markov chain, where your retort is only dependent on the last observed state.
Another fun topic was the application of the Shannon Entropy to text input on mobile devices via T9 and the different auto-correction algorithms around. Christian argues that these methods are shaping how you write, as they perform best on the vocabulary they were designed on and you thus unconsciously start to adapt to that vocabulary. Anecdotally this is just what I observe while typing away on my phone: Instead of the software adapting to my needs and figuring out what I wanted to say, I start to adapt to the idiosyncrasies of the auto-completion feature and try to predict what words the software can make most sense of, in order to avoid the software turning my intended writing into gibberish (especially for those üseless umlauts while typing in German. And I still have fond memories of T9 turning 'Sure, I'll come over and bring beer' into 'Sure, I'll come over and bring AIDS').
To borrow from Philipp's closings: Recommended for anyone interested in AI and computational linguistics (probably less if you're an expert in one of those fields).
January 27, 2012
Good, but oh, it could have been better.
There's an annual Turing test event in Britain every year. A group of top computer programs compete against a group of human confederates, as the computers try to prove, per Alan Turing, that they're really humans, just as the humans do.
So far, no computer has won this test, but, given the relatively narrow parameters of the test at this particular contest, that may not be too far off.
Christian, who successfully competed to be a human "confederate," takes off from that point in the paragraph above, to riff on what it means to be human (the human confederate the judges in the Loeber prize most frequently judge to be human wins "the most human human" award), larger issues in communication and information theory, and more.
Christian invokes the likes of Douglas Hofstadter at times, and in his last chapter, especially, does some Hofstadter-type pondering.
Contra others that gave this less than five stars, I didn't mind the digressive tone of the book at all; in fact, I loved it. Unfortunately, there's not a lot of depth or follow-through on the speculation.
This book could have been, and should have been 50 pages longer at a minimum. A full 100 pages of additional material, without getting as long as Goedel, Escher, Bach, or as technical, could have been doable.
Oh, and given the number of people mentioned in the book ... no index?
There's an annual Turing test event in Britain every year. A group of top computer programs compete against a group of human confederates, as the computers try to prove, per Alan Turing, that they're really humans, just as the humans do.
So far, no computer has won this test, but, given the relatively narrow parameters of the test at this particular contest, that may not be too far off.
Christian, who successfully competed to be a human "confederate," takes off from that point in the paragraph above, to riff on what it means to be human (the human confederate the judges in the Loeber prize most frequently judge to be human wins "the most human human" award), larger issues in communication and information theory, and more.
Christian invokes the likes of Douglas Hofstadter at times, and in his last chapter, especially, does some Hofstadter-type pondering.
Contra others that gave this less than five stars, I didn't mind the digressive tone of the book at all; in fact, I loved it. Unfortunately, there's not a lot of depth or follow-through on the speculation.
This book could have been, and should have been 50 pages longer at a minimum. A full 100 pages of additional material, without getting as long as Goedel, Escher, Bach, or as technical, could have been doable.
Oh, and given the number of people mentioned in the book ... no index?
February 15, 2023
This was a most unexpected delight and an absolute must in the age of ChatGPT. I remember learning about the Most Human Human award in a Quartz YouTube video by Brian Christian himself and instantly ordering the book. The book walks us through Brian Christan’s obsessive determination to win the Most Human Human award - stay tuned for the results at the end of the book ;) - but most importantly his meditations upon what makes us humans. This my geeky friends is a treasure that hits that optimum intersection between philosophy, technology and AI, neuroscience and human psychology. It is filled with corky references, disorganised in a very pleasant way and overall just a very human elaborate discussion on what it means to be human in the age of AI. Ofcourse, Brian Christian is also the author of his more acclaimed ‘Algorithms to Live By��� so you can pick it up with absolute assurance that this guy A. Knows what he’s talking about, B. Will not disappoint. His thesis put simply: “the story of the progression of technology doesn’t have to be a dehumanising or dispiriting one, quite as you will see, the contrary.”
I wrote in the first page: A very human book, from a very human human to other humans. There is no method to this kind of madness and perhaps, that is the fundamental exercise of being
The Ridiculous Loebner Prize
First of all, the fact that this competition exists is hilarious in and of itself. At least the Most Human Human part. The Most Human Computer aspect of the competition was a self fulfilling prophecy ever since Turing published his paper. But it gets even more ridiculous when we hear more about the founders, Hugh Loebner made his fortune out of selling disco rolled carpets and his co founder actually mistook his computer for a girlfriend in 2007 and was dating an AI for over four months.
What can humans really offer
In the introduction, Brian demonstrates how little uniqueness we really possess as humans by reducing the great human feat to THE SENTENCE OF: “THE HUMAN BEING, IS THE ONLY ANIMAL THAT XXXX”. Showing us how syntax based language wasn’t it/ the apes do it too, tool using wasn’t it / Octopuses use coconut shells, maths?/ calculators are better at it. He questions our obsessive pursuit with that question and why we constantly pit ourselves against our computers as he puts it:
“does the fact that computers are good at mathematics in some sense take away an arena of human activity, or does it free us from having to do a nonhuman activity, liberating us into a more human life”
Throughout the book, we will see that so much of our human experience comes down to the art of conversation, the art of connection and the depth of our human connections. Human intelligence takes on a very different shape once we put things in perspective and see that what we CAN’T programme is what truly makes us special and there isn’t much that we can’t programme, but there is a meta version of existence that can never be reduced to cold blooded machine and that META layer is where we prosper, I like to think of it as the subconscious.
On Being [favourite passage]
“For me, though, complacency — because it is a form of disengagement — is a whisker away from despair. I don’t want life to be “solved”; I don’t want it to be solvable. There is comfort in method: because we don’t always have to reinvent everything at every minute, and because our lives are similar enough to others’ lives, the present similar enough to the past that, for example, wisdom is possible. But wisdom that feels final rather than provisional, an ending rather than starting point, that doesn’t ultimately defer to an even larger mystery is deadening. I won’t have it. Perception gives us right solutions.”
On the soul: Takeaways from Chapter 2
In ‘The Migratory Soul’, Christian takes on a very interesting discussion that will for sure, make you think and most importantly and rest assured, you will read it with glee.
- Max Planck’s study found out that while chimps follow each others heads in conversation, we follow each other’s eyes
- Philosophers speculated on the hierarchy of the mind, the brain and the soul: Descartes located the soul to be in the centre of the brain. The ancient Egyptians who thought the brain was useless but literally jarred almost all other significant organs, thought it lied in the heart, which coincides with most cultures. Although in Persian, Urdu, Hindi and Zulu, the liver was the central organ for thinking and consciousness. Socrates believed that the soul transcends human flesh; Plato gives us the unified soul structure of “appetite, spirit and reason”; and Aristotle similarly breaks down the soul into three: an “appetitive soul” that comes from movement, “a nutritive soul that stems from biological nourishment” and a “rational soul” that was distinctive to humans and he was keen on the fact that the soul was an effect, “not the cause”. Stoics however placed the mind at the heart and the soul was a life force driving our psyche and mental states which plants didn’t possess. Descartes however believed there was one soul, which solely belonged to humans. Brian’s point of view is that “the soul body interface might be in a special place where normal, deterministic cause and effect physics break down, which is metaphysically awkward to conjure”
- The purpose of living according to Aristotle: “eudaimonia” which means happieness/ to flourish/ success/ the wellbeing of the spirit. Nietzsche however believes the purpose is to be and do good.
- Aristotle also attributes the uniqueness of humans to “reason”. Denoting that “contemplation” is what we do best. The stoics agreed but the Epicureans believed that “qualia” or the sensory experience of the soul from registering and experiencing the world was what we humans experience best but that was a reduced form of being for Stoics and Socrates. Descartes ofcourse said it best in his “Cogito ergo sum” until deductive logic, became AI’s mission.
- ‘Psyche’ was used in Homer’s text to refer to humans, but ever since it has taken a life of its own signifying some kind of “life force to describe anything from humans to greass”
- The left brain bias and dominance is attributed to many factors but mostly the fact that it is difficult for neurologists to discern/ map the right hemisphere whereas it is more easy to locate functions in the left side. When it comes to language processing for example, the left brain governs the spoken word but the right brain “concern itself with allegory, metaphysical and ambiguity”. The rise of the left in the way civilisation has been structured is due to the fact that its the side that “does all the talking”
- In more conservative communities where the emphasis of the human experience resides on contemplation and intellectual pursuit only usually results in a community where people LIVE AS MINDS. A relatable experience to me personally. Neglecting the body only leads to dissonance and lack of harmony, which is a claim we can use AGAINST ARISTOTLE’s notion that we are here to reason and contemplate and that’s all.
- The decision dilemma and an important reminder: what we really want is to be satisfied with our choices, that’s all, more information helps but doesn’t radically change our ability to make a decision; the hyper rational mind doesn’t help us evolve, it freaks us out
Overall
So basically the book is filled with anecdotes, exchanges and actual scripts between judges and chatbots/ natural linguist AI’s, dating advice, very human stories about very human people, back stories of corky nerds, real questions about what it means to just be human or OURSELVES, extremely insightful perspectives from the greatest thinkers of the world about all the domains in the philosophy of the mind, sudden but often entertaining POP references from anything from Sex and the city to Chocolat the movie to Hitchhiker’s guide to the galaxy to My Fair Lady and Ponochio, great book recommendations and so much more, you really won’t want to put it down.
Referenced Books to Pick Up:
Hitchhiker’s Guide to the Galaxy, Godel, Escher and Bach, Soul Made Flesh, Predictably Irrational, Zen and the art of motorcycle maintenance, Orwel’s ‘Politics and the English Language’, Mind Over Mood etc.
I wrote in the first page: A very human book, from a very human human to other humans. There is no method to this kind of madness and perhaps, that is the fundamental exercise of being
The Ridiculous Loebner Prize
First of all, the fact that this competition exists is hilarious in and of itself. At least the Most Human Human part. The Most Human Computer aspect of the competition was a self fulfilling prophecy ever since Turing published his paper. But it gets even more ridiculous when we hear more about the founders, Hugh Loebner made his fortune out of selling disco rolled carpets and his co founder actually mistook his computer for a girlfriend in 2007 and was dating an AI for over four months.
What can humans really offer
In the introduction, Brian demonstrates how little uniqueness we really possess as humans by reducing the great human feat to THE SENTENCE OF: “THE HUMAN BEING, IS THE ONLY ANIMAL THAT XXXX”. Showing us how syntax based language wasn’t it/ the apes do it too, tool using wasn’t it / Octopuses use coconut shells, maths?/ calculators are better at it. He questions our obsessive pursuit with that question and why we constantly pit ourselves against our computers as he puts it:
“does the fact that computers are good at mathematics in some sense take away an arena of human activity, or does it free us from having to do a nonhuman activity, liberating us into a more human life”
Throughout the book, we will see that so much of our human experience comes down to the art of conversation, the art of connection and the depth of our human connections. Human intelligence takes on a very different shape once we put things in perspective and see that what we CAN’T programme is what truly makes us special and there isn’t much that we can’t programme, but there is a meta version of existence that can never be reduced to cold blooded machine and that META layer is where we prosper, I like to think of it as the subconscious.
On Being [favourite passage]
“For me, though, complacency — because it is a form of disengagement — is a whisker away from despair. I don’t want life to be “solved”; I don’t want it to be solvable. There is comfort in method: because we don’t always have to reinvent everything at every minute, and because our lives are similar enough to others’ lives, the present similar enough to the past that, for example, wisdom is possible. But wisdom that feels final rather than provisional, an ending rather than starting point, that doesn’t ultimately defer to an even larger mystery is deadening. I won’t have it. Perception gives us right solutions.”
On the soul: Takeaways from Chapter 2
In ‘The Migratory Soul’, Christian takes on a very interesting discussion that will for sure, make you think and most importantly and rest assured, you will read it with glee.
- Max Planck’s study found out that while chimps follow each others heads in conversation, we follow each other’s eyes
- Philosophers speculated on the hierarchy of the mind, the brain and the soul: Descartes located the soul to be in the centre of the brain. The ancient Egyptians who thought the brain was useless but literally jarred almost all other significant organs, thought it lied in the heart, which coincides with most cultures. Although in Persian, Urdu, Hindi and Zulu, the liver was the central organ for thinking and consciousness. Socrates believed that the soul transcends human flesh; Plato gives us the unified soul structure of “appetite, spirit and reason”; and Aristotle similarly breaks down the soul into three: an “appetitive soul” that comes from movement, “a nutritive soul that stems from biological nourishment” and a “rational soul” that was distinctive to humans and he was keen on the fact that the soul was an effect, “not the cause”. Stoics however placed the mind at the heart and the soul was a life force driving our psyche and mental states which plants didn’t possess. Descartes however believed there was one soul, which solely belonged to humans. Brian’s point of view is that “the soul body interface might be in a special place where normal, deterministic cause and effect physics break down, which is metaphysically awkward to conjure”
- The purpose of living according to Aristotle: “eudaimonia” which means happieness/ to flourish/ success/ the wellbeing of the spirit. Nietzsche however believes the purpose is to be and do good.
- Aristotle also attributes the uniqueness of humans to “reason”. Denoting that “contemplation” is what we do best. The stoics agreed but the Epicureans believed that “qualia” or the sensory experience of the soul from registering and experiencing the world was what we humans experience best but that was a reduced form of being for Stoics and Socrates. Descartes ofcourse said it best in his “Cogito ergo sum” until deductive logic, became AI’s mission.
- ‘Psyche’ was used in Homer’s text to refer to humans, but ever since it has taken a life of its own signifying some kind of “life force to describe anything from humans to greass”
- The left brain bias and dominance is attributed to many factors but mostly the fact that it is difficult for neurologists to discern/ map the right hemisphere whereas it is more easy to locate functions in the left side. When it comes to language processing for example, the left brain governs the spoken word but the right brain “concern itself with allegory, metaphysical and ambiguity”. The rise of the left in the way civilisation has been structured is due to the fact that its the side that “does all the talking”
- In more conservative communities where the emphasis of the human experience resides on contemplation and intellectual pursuit only usually results in a community where people LIVE AS MINDS. A relatable experience to me personally. Neglecting the body only leads to dissonance and lack of harmony, which is a claim we can use AGAINST ARISTOTLE’s notion that we are here to reason and contemplate and that’s all.
- The decision dilemma and an important reminder: what we really want is to be satisfied with our choices, that’s all, more information helps but doesn’t radically change our ability to make a decision; the hyper rational mind doesn’t help us evolve, it freaks us out
Overall
So basically the book is filled with anecdotes, exchanges and actual scripts between judges and chatbots/ natural linguist AI’s, dating advice, very human stories about very human people, back stories of corky nerds, real questions about what it means to just be human or OURSELVES, extremely insightful perspectives from the greatest thinkers of the world about all the domains in the philosophy of the mind, sudden but often entertaining POP references from anything from Sex and the city to Chocolat the movie to Hitchhiker’s guide to the galaxy to My Fair Lady and Ponochio, great book recommendations and so much more, you really won’t want to put it down.
Referenced Books to Pick Up:
Hitchhiker’s Guide to the Galaxy, Godel, Escher and Bach, Soul Made Flesh, Predictably Irrational, Zen and the art of motorcycle maintenance, Orwel’s ‘Politics and the English Language’, Mind Over Mood etc.
June 19, 2011
The context for this exploration of human intelligence is an annual contest in which judges engage in casual 'chats' with computers and humans and try to determine which is which. The point of the contest, of course, is to determine the "most human" computer, a modified Turing test. The risk to the human entrants is that they might be judged a computer! The author was determined not to let that happen to him and so he engages in one of the most entertaining explorations of what it means to engage in distinctively 'human' conversation. One flaw perhaps is that there is relatively little here about the details of the conversations that convinced the judges that the author (spoiler alert)
was not a computer, or to show whether the computers were particularly inept that year at seeming human. Still, there are bits of transcripts from years of the contest to illustrate the strategies
that do and don't work.
Recently, I've read about a man who trained his mind to remember random information, but not where he parked his car (Moonwalking with Einstein), and a woman who obsesses over every detail of her life and each day's events, but can't remember phone numbers (The Woman Who Can't Forget). But this book convinces me that it would be far better to have some imagination and the ability to tell a good story than to have either of these extraordinary memory talents.
was not a computer, or to show whether the computers were particularly inept that year at seeming human. Still, there are bits of transcripts from years of the contest to illustrate the strategies
that do and don't work.
Recently, I've read about a man who trained his mind to remember random information, but not where he parked his car (Moonwalking with Einstein), and a woman who obsesses over every detail of her life and each day's events, but can't remember phone numbers (The Woman Who Can't Forget). But this book convinces me that it would be far better to have some imagination and the ability to tell a good story than to have either of these extraordinary memory talents.
June 3, 2013
I loved this book. I started with rather low expectations; sort of expecting a new york times style high level view of machine learning and its implications. What I got instead was profound insight into what machine intelligence can tell us about what it means to be human. I suspect I'll look at life differently, both my human interactions and my observations, for weeks to come. I wish I'd been taking notes, because every time there was a great concept which I thought deserved more thought or research, another one would come along and replace it. I'll definitely have to reread this book. If you're interested in philosophy or in how stuff works, I can't recommend this book enough.
November 16, 2011
A very interesting, well researched book on what it means to be human in light of developments in artificial intelligence. The traditional factors that defined what it was to be human no longer apply due to the fact that computers can exhibit most, if not all, of these qualities - especially reason - so where does that leave us? What does set us apart? This book is highly thought-provoking. The author does wander quite a bit and it could have been more cohesive but overall it was very enjoyable.
October 26, 2019
Christian writes from the perspective that it will be bad if and when machines are intelligent, he's kind of an anti Ray Kurzweil. I completely disagree with his viewpoint but his book is interesting and illustrates lots of Turing problems that I'd never considered.
He goes on to say that it will not happen that the same bot will win year after year but is proved wrong just 2 years after the book is written by Mitsuku who won in 2013,2016-2019
https://en.wikipedia.org/wiki/Mitsuku
He goes on to say that it will not happen that the same bot will win year after year but is proved wrong just 2 years after the book is written by Mitsuku who won in 2013,2016-2019
https://en.wikipedia.org/wiki/Mitsuku
September 16, 2015
An exceptional book on what it means to be truly human, truly empowered for uniqueness and collaboration.
Also, as literature, so many precise words and exquisite constructions. Stretched my mind.
Also, as literature, so many precise words and exquisite constructions. Stretched my mind.
Read
February 13, 2024Ah, dang, waited just a few too many years to read this book and now it's wildly obsolete. I guess it functions as a good historical document of the dark ages of chatbots before LLMs and most of it is really about random little quirks and features of humans and related issues, so not a total loss, but I should really do a better job of getting through my to-read list in a timely manner.
Want to read
November 10, 2022books referenced or recommended in the books that i am currently reading >>>
August 3, 2016
My favourite trait of Most Human Human was its optimistic tone. I have read a fair amount of AI criticism, lamenting that it might be very dangerous, moaning about it taking jobs away from people and whining that it is making us more stupid or at least less intelligent. Instead, this book embraces human endeavours directed into this field, highlights the achievements and their implications and treasures the foreseeable future of AI. The author has a very good reason for this, too: artificial intelligence has made us reappraise what it means to be intelligent. It has made us reconsider what it is to be human. Also, Brian Christian suggests that instead of the last few decades being a race of machines against humans, we are in it with them. We need to learn to co-exist and since we already have a pretty good idea regarding what computers excel at, we need to find out - or maybe rather rediscover - what we our qualities are. What makes humans... human.
Or it least that's how I interpret it. The book will walk you through the basics of chatbot design and development and take you on the bumpy ride through the recent history of AI and computers in general. It will leave you ruminating on the art of conversation, not leaving out speed dating, pickup artists or the fact that heated arguments are non-contextual in the sense that the answer typically depends solely on the previous question. You will meet ELISA, a psychotherapeutic bot that, basically, by rephrasing questions, provides people with real therapeutic experience, leaving them feeling jollier afterwards.
I really enjoyed the section dealing with computer chess and a different one focusing on entropy and statistics in the English language. The whole read is like this: a bit all over the place, pulling out information from every field that might add to the topic. And since the topic is humanity, there is a decent number of potential candidates. Apart from solid grasp of IT, the author seems to be educated (to an extent, anyway) in philosophy, poetry and music among other fields. The author was also 24 or something when he wrote this and he has my eternal utter respect for this achievement.
It's about art, literature, small talks and about the fact that only real conversations, i.e. the ones where you cannot be finishing the other person's sentences, matter. The thing is full of beautiful quotes but one that stuck in my mind is that maybe, it is better to be asking: "How did you stay together?" instead of "How did you meet?". A lot of fairy-tales deal with the latter but almost none with the former.
Btw, a friend of mine, also a software developer, recommended this book to me. If someone has read this and you have no connection to software development in your everyday life, could you let me know whether you liked it or not? To me, it seemed that the book is written and organized in such a way that it must appeal to the way minds of people bossing computers around all day long typically function. I am just wondering how well suits anyone else.
Or it least that's how I interpret it. The book will walk you through the basics of chatbot design and development and take you on the bumpy ride through the recent history of AI and computers in general. It will leave you ruminating on the art of conversation, not leaving out speed dating, pickup artists or the fact that heated arguments are non-contextual in the sense that the answer typically depends solely on the previous question. You will meet ELISA, a psychotherapeutic bot that, basically, by rephrasing questions, provides people with real therapeutic experience, leaving them feeling jollier afterwards.
I really enjoyed the section dealing with computer chess and a different one focusing on entropy and statistics in the English language. The whole read is like this: a bit all over the place, pulling out information from every field that might add to the topic. And since the topic is humanity, there is a decent number of potential candidates. Apart from solid grasp of IT, the author seems to be educated (to an extent, anyway) in philosophy, poetry and music among other fields. The author was also 24 or something when he wrote this and he has my eternal utter respect for this achievement.
It's about art, literature, small talks and about the fact that only real conversations, i.e. the ones where you cannot be finishing the other person's sentences, matter. The thing is full of beautiful quotes but one that stuck in my mind is that maybe, it is better to be asking: "How did you stay together?" instead of "How did you meet?". A lot of fairy-tales deal with the latter but almost none with the former.
Btw, a friend of mine, also a software developer, recommended this book to me. If someone has read this and you have no connection to software development in your everyday life, could you let me know whether you liked it or not? To me, it seemed that the book is written and organized in such a way that it must appeal to the way minds of people bossing computers around all day long typically function. I am just wondering how well suits anyone else.
January 13, 2019
A lot of interesting data and theories shared about AI and perspectives we forget about as we just keep using tech every day.
Displaying 1 - 30 of 478 reviews