Skip to main content

tv   George Gilder Gaming AI  CSPAN  November 25, 2021 9:00am-10:02am EST

9:00 am
you can follow along behind the scenes on social media at book tv on twitter, instagram and facebook. some people say that artificial intelligence is going to make the human race obsolete. a lot of people don't want to think about ai, artificial intelligence. the thing about ai is, even if you don't want to think about it, it's thinking about you. or is it? that will be the question we will discuss today on independent conversations. greetings everybody who joined us. we try and bring experts to
9:01 am
discuss topics of the day and give you a perspective you will not likely hear. today, we will talk with george guilder. >> great to be here. >> it's a pleasure to see you again. i met you first -- i think it was in the winter, maybe january of 1982 in western new york. you had recently published "wealth and poverty," the year before. was it published in 1981? >> 1980. >> okay. it was a fascinating book. i think president reagan loved the book, if i remember hearing he read it. did you hear that story? >> he wrote me letters about it. before publication, he read articles -- excerpts from it.
9:02 am
it was excerpted all over the place before it came out. it made me president reagan's most quoted living author. >> it was a fabulous book. your creative in seeing what others didn't see about the system of free exchange, so-called capitalism, when you put -- >> the whole bunch of different ways that -- >> tribes would get together and give and share, which was fascinating. you pointed out that there's a lot of that in what we call capitalism, which, therefore, doesn't pivot simply on self-interest but rather on something akin to be a -- be ne
9:03 am
-- benevolence. thank you. >> my books sprang from "wealth and poverty" which focused on creativity in the image of our creator as the great force in economic growth. since then, i've been working on the information theory about economics. >> i remembered the term. i was trying to think of it a moment ago. you described it as the potlatch. >> yeah. >> that was really amazing. it helped me. i was a college student at the time or just after being a college student. i was having a lot of tussles with professors and peers who thought socialism was the coolest thing there was. they usually portrayed capitalism in distorted terms. you gave me a new vocabulary. >> thank you. >> people said you are an
9:04 am
economist. sometimes you seem like you are a sociologyist. other people say you are a few futurist. what are you? >> i'm a historian. i'm willing to play the role imposed. >> we're glad. >> i really probably believe in a higher -- it unifies all different fields. allows you to transcend the fragmentation of analysis that
9:05 am
afflicts all the universities where everybody has his own specialization. many of them with different jargon and idioms and expressions that exacerbate fragmentation of knowledge. >> they do. you work is integration rather than fragmentation, which makes sense. that's probably what has driven you to be part of the co-founders of the discovery institute. they seem to have quite an understanding of the sciences there. >> that's what we try to do. we try to bring the sciences together. economics is just another part of biology, which is another part of physics, which is in a cosmic vision that we are going
9:06 am
to expound on in our conference november 10 to 12. peter thiel will be the keynote. we have bob metcalf will be expounding on the continued significance for cryptocurrency and other such paths of ecological advance. we will have an exciting time. i'm going to debate newt gingrich on china. >> oh, ho, ho. >> i don't think war with china brings any benefits that i can imagine. >> i agree with you on that. what could be less productive than a war with china? good grief. if our viewers want information about that conference, where should they go? >> cosm.technology.
9:07 am
>> go there to find out about this conference that's going on. next month did you say or november? >> november. november 10 to 12. >> in the meanwhile, you are releasing a new book. i think the publication date is officially october 15th, if i'm not mistaken. here is the cover of it.
9:08 am
>> let's talk about the book. i got a copy of it. i was fascinated by the way that you take up the standard challenge and turn it in a direction that people don't expect. the standard challenge -- you mentioned early in the book that some people think that ai is going to be for sure a demotion of the human race. i think on page 20 of the book -- it's a very arresting quote. caught my eye where you quoted the late steven hawking who pronounced the development of full artificial intelligence could spell the end of the human race. >> that's what hawking said. elon musk, who is alive today, says ai is more dangerous than nukes. a lot of people talk about sing
9:09 am
singularity to come. this was predicted way back by alan turing's colleague, jack good. he said once we invent artificial intelligence, that will be the last invention we will ever have to make. true artificial intelligence would be capable of creating machines -- intelligent machines that can outperform the original artificial intelligence, and release a cascade of intelligence through the universe. >> culminating -- theory was, it was culminate in this so-called
9:10 am
singularity where the artificial intelligence takes up where we left off and says good-bye to us. right? >> yeah. that's really what good was predicting. a lot of people took it further. it responds to email. it has responses that allows you
9:11 am
to anticipate how you are going to respond to a particular email. >> i notice those responses have been getting more courteous and more specific. i suppose it's due to his development. >> that's his contribution. i think all these people have forgotten a fundamental principal of the computer science. >> that's what's striking about this book. you don't seem to be as much of a doom sayer. you seem to think that the potential of ai may be oversold, but that even in overselling, there could be collateral damage. you are trying to avoid that. have i got that right? >> yeah, i think that's right. the idea that somehow ai competes with human minds is a
9:12 am
fundamental illusion. >> a lot of these technology creators, they came to their work having already absorbed the idea that the human mind is nothing more than a meat machine. if they knew that to begin with, then it's not surprising that their conception of artificial intelligence could be the singularity thing. if the human mind was nothing more than meat and electrons, then you could surpass it. your point is that the human mind demonstrates that it must be more than just meat and electrons. >> when i was writing about the internet, which i did from the late '80s on through its development and the launch of
9:13 am
the webs around the globe, i used to -- a way of mapping all the connections of the global internet. about two data bytes. >> reminder what of zettabyte is. >> ten to the 21. 21 zeros and just the number beyond easy a number beyond imagination. recently, mit campaign has been
9:14 am
trying to map all the connections in a single human brain. >> they start with -- a friend of mine was on the team that developed dna codes. imagine that dna was a code and worked out what the code would be. he has been mapping brain of a worm for 20 years at the university of wisconsin. at thanksgiving dinner the other year, last year or so, he told
9:15 am
me that the more he studies, the less he understands the brain. the folks at mit have taken his connections and applied it to the brain of a human being, with all the connections and all kinds in the brain. it turns out, it takes a lot to map all the connections. it suggests a single human brain is as densely and complexly connected as the global internet is.
9:16 am
the global internet takes gigi -- gigawatts of energy. it is next to a glacier to deal with heat problems. generally, the dominant technology of a data center is all the cooling systems to take away the heat that these machines emit. single human brain functions with 12 to 14. >> i'm just 98.6 degrees celsius. >> i believe that technology
9:17 am
functions to the extent that it extends our human capabilities rather than attempting to compete with human capabilities, let alone usurp human capabilities. companies in silicon valley that regard their business plan as obsolete and their customers and contributors are going to fail. >> if that's how they approach it, then they're going to make themselves superfluous. you will run out of things to do. >> i think it's absurd.
9:18 am
i think that the nobel laureate economist william nordhouse did a study of the advance of -- the invention and creation of white, the amount of lumen to light a room at night. he shows that the advance in lighting has been 100,000 times more rapid than is measured in the economic model.
9:19 am
the incredible expansion of light from the time that was -- it was fire in a cave to the millions of candles at versailles to whale oil to kerosene to finally electricity and then light emitting -- >> beyond electricity. >> measured by the amount of time a worker had to spend to buy the light to illuminate a room. economic progress was 100,000
9:20 am
times more rapid than is usually estimated in that field. during the industrial revolution, we missed light. this continues to be, a golden age of capitalism with the technological progress as fast
9:21 am
as ever and with increasing the quality. because poor people benefit more from the expansion of the hours of their day to do other things than rich people already just have to spend a few minutes to earn their food and clothing and whatever. and so as technology advances, it benefits the masses most, and ai is just the newest manifestation of the advances of the computer industry since the time of touring and good at bletchley park through john von norman who was probably a paramount figure and anticipated
9:22 am
the gigahertz machines we have today. he really was the first person to imagine moore's law, that we could produce making the machines that operators at billions of cycles a second. >> let me step back a couple steps to something said a minute ago, really deserves i think extra attention. you commented a moment ago that technological and economic advance tends to have a comparatively greater impact the benefit to the worse off, because the the worst off tt further to go up. and so the comparative improvement in their life can be greater. that's intriguing. a few years ago i was in east africa in uganda and traveling around some of the rural areas and a part of uganda. it was striking to me of course
9:23 am
the standard of living obviously much lower than the united states and and i saw many pe living in hats, not having sufficient clothing, not having sufficient covering from the rain. -- huts. people clearly struggling although there was a lot of economic activity and at the same time every single person sitting under every insufficiently corrugated tin roof on every little shop when every little byway or alleyway have a cell phone. every single person has a cell phone. >> and increasingly it's a smart phone. >> and they use it -- >> that means a a supercomput. that means an underestimation of the real standard of living. >> that's right. >> the same kind of factor of a hundred thousand that nor house identified during the industrial revolution with expansion of light. it's another form of the
9:24 am
expansion of light. >> i notice ugandan roads were still pretty bad and need to be repaid at the same time i also realized that everybody can now talk to their grandma or great grandma out in the backcountry anytime they want to. because everyone even the small villages as cell phones and also using the smartphones as a meeting of payment and exchange greatly supplying monetary transactions. it was quite stunning honestly and made me proud to be a northern california. >> you are correct to be proud and there's really what's bizarre is the argument that you see a lot of places that the middle class is suffering as a result of stagnation of technology or whatever is the claim of the moment, that
9:25 am
inequality is vastly expanding. you know, once you have, if you score a thousand dollars, that takes care of all your essential needs and you live a lot better than a king of previous -- if you have that smart phone and access to medical care, that it implies and ultimately the access to a whole world civilization, that it manifests. rich people, all their so-called wealth, is really knowledge. it's invested. it's not liquid.
9:26 am
>> right. >> ultimately it disappears. in capitalism you only get to keep what you give away. that is because, because unless your wealth is invested and is working and providing jobs and opportunities for others, it loses value and ultimately disappears. so this is really a fundamental principle of capitalism, and it's manifested today in the phenomenal creativity that you saw in uganda. >> we have a number of people on with us, george, simultaneously although we also may share this recording later but one of her current participants sent a note in commenting that an organization or company called
9:27 am
solaris technologies in san jose is a good example of the kind of thing you're talking about. do you know solaris technologies by any chance? >> how do you spell it? i have heard of solaris technology. i was thinking of soul brass which is i think the more formidable accomplishment. it's a wafer scale integration of ai and a shared learning capabilities on the single-chip not the size of your thumbnail but of the size of a dinner plate, and trillions of transistors on a single wafer. i can't remember what the heck
9:28 am
so lera does. >> something good apparently. >> he should tell you, if are going to talk about it, he should tell us which company that is. >> i'm watching the comment box. we'll see but one of the great arguments in the book, "gaming ai," is your point that those in the high-tech industries who are, obsess maybe or maybe captivate is a nice a word with his idea of movement towards a a single dirty weather created intelligence surpasses human mind and so forth and makes the human mind obsolete. they seem, you argue, to africa on the history of their own industry. >> that's right. >> can you tell me something about that in a way that i as a layman can understand? how does the development of high-tech it industry itself illustrate your point about the irreducible need for the creativity of the human mind?
9:29 am
can you tell me something about that? >> well, i brought up john van norman, the great figure who imagine that you could make mathematics completely self-sufficient. , rendered mathematics a completely self-sufficient and complete and coherent system. he met this young student named kurt gödel who i believe was the inventor of the computer science of our day. >> kurt in 1931, i believe, introduced the paper which
9:30 am
showed that mathematics was intrinsically dependent on axiums or propositions that couldn't be proven within the system itself. >> couldn't be fully self-contained. >> it could not be a self-contained system. the greatest mind that we have produced over the last couple centuries, vonn noyman concluded from this, he was the only one who understood the paper, and he not only saw that this meant that thinking machines would always be necessarily dependent on outside programers or
9:31 am
oracles. >> we understood the
9:32 am
artificial intelligence. it was a massive expression of the disabilities of human lives extended into the world class it actually is an extension, not a replacement. what you said a moment ago really is a way of capturing it. tell me again who it was that made the point that all this developed machine intelligence would have to have a human mind as if it were an oracle. who was said the oracle business? >> that was alan turing. >> it's a very striking metaphor because it means to put it in simple but historical terms, we are arguing machine intelligence.
9:33 am
what the oracle of delphi was thought to be to the men of antiquity. in other words a form of knowledge mysteriously outside the realm of grasping so you would go to the oracle and it probably was a bunch of who we but nonetheless they would go there and they thought they were receiving insights which they couldn't get with the human mind so the human mind in terms has turned to the artificial intelligence the way the oracle was to humansof that day. it's a fascinating metaphor . >> and it ultimately is an extension of charles sanders proposition that all information is triadic. it can't be binary.
9:34 am
if it's binary it's restricted to simple systems. and there's no necessary connection between simple as in mathematics and objects such as objects of those worlds. in order to connect the simple systems of the real world need an intermediating mind. human consciousness, human brain. >> it's sort of like two dimensions versus three dimensions . >> that's right. >> binary versus triadic. >> and we have a flat universe society prevailing in silicon valley. they imagine that there binary symptom symbols can play games a lot better than
9:35 am
us because they came in the symbol system so there's on the billboard there is just those black and gray, white andgray stones . those stones are symbols and they don't point beyond the board. if the computer can move those stones billions of times faster than a human can , obviously they can play go better than the human just as threshing machines can ... >> your point is a man threshing is going to be superseded by a threshing machine . that doesn't mean thatsomehow the threshing machine is more sophisticated than the man . >> that's right. >> you say early on in the book and you repeat it a few different places that youhave two basic claims . this notion of sort of supremacy of artificial intelligence you say is both
9:36 am
done and self-defeating so we've been dealing with the dumb part . i find it reassuring to learn from you in your book and other sources that the human mind actually is more complex than maybe the entire world and that's reassuring. i'm glad to know maybe there's evidence that i'm reminded is more than just a meat machine with electrons pulsing through it. but that's reassuring. maybe it's dumb but the trouble is that this view of artificial intelligence rising to a supremacy over everything, could it be self-defeating? i get that it's mistaken but how could it beself-defeating ? how could it undermine to >> they could have tried to replace the ccustomers.
9:37 am
the computer technology they are creating, an expression of their genius and human imagination and ability to have counterfactual projection and to imagine what doesn't already exist or is not already in the program. >> that's the human part. right? >> that's the human part. why the medical -- everybody decades ago was -- a doctor invented a diagnostic machine using a mainframe or maybe a digital mini-computer. the idea was that inevitably,
9:38 am
the machine learning or artificial intelligence would excel all human diagnosticians. once the symbols are prepared, once you store all the input for the machine and get them all categorized and tagged correctly, then an algorithm can function as billions of cycles a second and produce an answer. much of the intelligence is communication between the textures of the real world and the symbols that express it within the machine. we now have the illusion of
9:39 am
quantum -- i wrote a book about quantum -- >> what was the title? >> "microco"microcosm." that published in 1989. it was called the quantum era in economics and technology. of course, it's the quantum era. the whole semiconductor industry and its history was based on manipulating matter from the inside in accordance with quantum principles. all technology, all our computer technology is based on quantum physics. quantum fizz sikz the theory of the microcosm, whatever.
9:40 am
the problem is connecting these -- the system to the real world. what they call quantum computer does is abandon the binary on/off switches that have been the salvation of computers and use q-bits which are more complex analog systems. quantum computing is really a return to analog computing. analog computing was displaced by digital computing, not because analog computing wasn't
9:41 am
faster and it didn't correspond more closely with the real world but because in analog computing, making a -- analog model of the world takes endless details, map ing of the real territories and textures of our actual existence onto the computer. so analog computing -- quantum computing is terrific, but it imposes the whole burden on the human minds that program it. the problem gets moved from the symbol realm into the analog
9:42 am
realm where it incurs all the complexities and quantum uncertainties that populate the quantum world. >> the human mind -- again, layman's term here. human mind can set up closed systems which can then maybe run artificially better than the human brain could run them. that could imagine and create systems that are outside of the closed system. the human mind seems to be able to transcend closed systems and introduce new angles. that's what generates and powers creativity. if those people in charge of these industries deprecate the role of human mind and creativity, they may end up putting their own enterprise on
9:43 am
the road to if not failure at least less creative. is that right? >> i think that's beautifully stated. i think that creativity always comes as a surprise to us. >> i hope that your colleagues in this industry over across the bay here in silicon valley, which is not far from where we are at the independent institute on the other side of san francisco bay, i hope that they pay attention to you. if not, and if you are right, it might be that they will be overtaken in creativity, because they will be deprecating the very qualities that made their business work, which seems like it would be a terrible shame. pay attention to george. >> one thing, they can pay attention to the history of their own industry. >> there you go. >> pay attention to the
9:44 am
universe, this idea that the human mind is a product of random fluctuations of molecules is delusion to begin with. it's this belief that the human mind is the product of random evolutionary forces that really -- it makes them think that they can duplicate their minds with a machine. the mind is almost infinitely more complex than the machine they are building even today. >> to those of us -- >> they don't understand it at
9:45 am
all. >> our friends of the creative technological enterprise, we would encourage our colleagues, as it were in creativity, not to underestimate their own minds by buying into this really ridiculously reductive idea that the mind is nothing but a random set of physical mechanics. >> they have to move to life after google. >> this is your immediately life after google. i'm going to take you somewhere unexpected here, i think. partly because i'm looking at some messages coming from our viewers right now. following up what you said about the effects of this belief, this faith that the mind is nothing other than accidental material, mechanical and physical, this person says, the simplest of mind and the simplest of persons
9:46 am
is more complex than the world internet system. it's hard to say there's no god. that's what this person says. i think that's a good point. i have another point that's a little different from that, which is this. let me try this on you, george. i was reading your book, this thought struck me, which is that there is always has been in the history of our civilization something of a tension, if not always opposition, between a mindset which is empirical in nature and a mindset which is spiritual and pious in nature. that's why people say religion and science have been in each other's way. there's something to that. what i'm seeing based upon your analysis, is that religion is the religion of singularity
9:47 am
which is in the way of creativity. this is a replay in a really unexpected form of an old opposition. or is it? i'm thinking it's a replay of a very old opposition but the roles are reversed, because people who are all gaga over the power of ai to take over everything in the form of singularity, they are so committed to their faith position that they seem to close off their ability to be receptive to other data. you are bringing other data in. you are the scientist. they are the priest. >> one of the inventors of virtual reality, did the first virtual reality machines, there's been a number of good books on this subject. he says ai makes you stupid essentially. >> that's your point. very intriguing. here is another comment from one
9:48 am
of our viewers on with us right now. this person -- jennifer says, let's hear something about ai and its military implications. drone technology and the ability to select targets without human interaction, for example. do you know enough about this to comment on ai and military application? >> computer -- our whole military is based on computer systems. manhattan project was all modelled on computer systems where richard fineman got really immersed in computation as part of the manhattan project. fineman makes the crucial
9:49 am
observation that when you are building technology, you better respond to reality, because reality can't be fooled. the reality is that these machine learning systems are completely dependent on human minds. they do not think at all. the idea that these machines are somehow thinking as they shuffle bits and bytes is a religious belief. it's a particularly stultifying religion. >> that was my point. that actually means there's national security danger in deploying artificial intelligence on the assumption that it can think for itself. >> that would certainly be true. i don't think they are quite
9:50 am
doing that yet. but they are advancing drones probably too quickly. they probably are exaggerating their capabilities. >> learned recently in kabul. one yeah, one of them didn't work the way that president biden thought it was going to work and that was pretty disturbing. >> and children from military group. >> yeah with well that's disturbing and moreover. >> i'm not debunking ai. i think ai is great. just another step in the e evolution of computer industry. it poses no threat to human beings. >> right. >> i mean, the idea that it is comparable to nukes as elon musk
9:51 am
describes it is true only that nukes can be deployed by human minds and ai can be used to deploy nukes. but it is the human mind and the law and order and the civilization that keeps this alive. and if we imagine that our whole civilization is a product of random mutations of chemistry and physics. i think that is the flat universe theory. >> physics and chemistry. and that is stultifying and ultimately disabling philosophy.
9:52 am
>> we spoke a minute ago of the failed drone strike in kabul that killed a family with children. and you said it wasn't able to distinguish. okay. but what's striking to me about that example is that if drones were made more sophisticated by their human creators, they might be able to make such distinctions, at least approximate them better. but the problem is what if the creators of the drone artificial intelligence themselves don't think that human beings are anything special? they don't necessarily believe there is something special about mothers and children. what if they don't believe that and they are the ones creating the artificial intelligence to run the drones? that scarce -- >> -- well, all of -- order. >> right. >> and creativity in the image
9:53 am
of our creator. and that's the foundation of human life and progress. and it is baseballed and crippled by a conception that somehow we're just machines and our machines can adequately replace us. >> the understanding that our creativity because we bear the creator god is not an obstacle but maybe the -- >> -- i agree with that propagation. proposition. >> i'm thinking of that great book "the savior of science" published in the 1960s, which made that point before all of this began to happen.
9:54 am
we've got another interesting comment. we're going stop soon here but another interesting comment. one of our participant, laura, actually a friend of mine i think. wrote in saying, can moral or ethical checks and balances be programmed into ai? >> it implies that ai has consciousness, the potentiality for conscious. what it has is a program. and you can program any constraints that you want in the machines that you build. and you have to do it. and i mean, so but not as if you
9:55 am
are programming a conscienceness. but a series of programs and parameters.consciousness. but a series of programs and parameters. >> jacob wrote in saying i would like george to provide his insights into the future what the world will look like in ten years, thirty years. >> i expect the domination of the theory of -- of the information theory of economics, which prohibits really anticipating the future. the future is based on human creativity. and princeton hirschman declared, creativity always
9:56 am
comes as a surprise to us. and no deterimistic theory of economics, no deterimistic theory of lines can create a new future. and what differentiates our age from the stone age is not a steady refinement of stones. it is in advance of knowledge. knowledge is wealth. growth is learning. and it is all constrained by the passage of time, which is what remains scarce when all else grows abundant. so the future is not just going to be more of the same, in other
9:57 am
words degeneration. its got to surprise us. and i believe that in 30 years, we're going to live in a world that is -- would be almost incomprehensible in some ways. technologically from the world we live in today. i think we'll go beyond silicon. i think we'll produce and our intelligent machines will depend on a new carbon age. that just as our brains are consist of carbon, so will our intelligent machines of the future consist of various forms of carbon. they're already getting introduced in the form of carbon nano tubes.
9:58 am
graphene devices and other new hybrid materials that -- that can simulate intelligence better than our silicon, binary silicon machines of today. so i think what, we'll have a life of silicon. >> i think carbon -- isn't carbon more plentiful element than -- >> no. it's less plentiful but it's more -- the three -- silicon is great because it is one of the three most common elements in the earth's crust. which gordon moore, the intel founder believed was approved
9:59 am
providencial. i believe the new substrates of the new intelligent machines will be carbon-based. >> at the end of the book "gaming ai," george you say these interesting words. and i they'll stop here. you say an explosion of productivity does not mean an evaporation of work. ai will make -- thus more employable. create new and safer and more interesting work. it will generate the capital to endow new companies and new ventures as new technologies have done through history. what it will not do is create a mind. three cheers for the human mind
10:00 am
and what it tells us about the universe. thank you george gilder. so grateful for you taking the time. thank you for writing the book "gaming ai." thank you for writing earlier book "life after google" and thank you physically for being a friend of us. >> -- >> we refer friends to our friends up in seattle at the discovery institute. and again, thanks to george gilder and thanks to everybody who joined us for today eats independent conversation from the independent institute here in oakland, california. have a great day and please join us again. thanks george. bye-bye.
10:01 am
afterwards is a weekly interview program with relevant guest hosts. interviewing top authors about their latest work. >> the themes are relevant for today and the pandemic. i know you said you started writing the book around the time the pandemic began. but can you tell us where the idea for the book came from and why this book now? >> i think the emotion for this book actually came from

13 Views

info Stream Only

Uploaded by TV Archive on