Home Videos

Our AI Future Is WAY WORSE Than You Think | Yuval Noah Harari (YouTube Video Transcript)

Need transcripts for other videos? Try our YouTube Transcript Generator →
Title: Our AI Future Is WAY WORSE Than You Think | Yuval Noah Harari
Duration: 01:37:44
Total Correct Answers:
Current Caption
Correct

Learning Modes

YouTube Video Transcript Hide

Ask AI Result

The ask AI result will appear here..
(00:00:00) Your YouTube transcript will appear here (00:00:01) most people around the world are still (00:00:03) not aware of what is happening on the AI (00:00:06) front it can invent medicines and (00:00:09) treatments we never thought about but it (00:00:11) can also invent weapons that go beyond (00:00:14) our imagination you're changing the (00:00:16) basis of everything it's no wonder there (00:00:19) is an earthquake in the structure that (00:00:21) is built on top of it I got news for you (00:00:24) people the rise of the Machines is (00:00:27) already upon us so what exactly do we (00:00:30) need to understand about the rapid (00:00:32) Ascent of artificial intelligence what (00:00:35) does this revolution augur for the (00:00:37) future of the human species to gain (00:00:39) Clarity amidst the confusion I'm joined (00:00:42) today by Yuval Noah Harari a (00:00:44) world-renowned historian and mega (00:00:47) best-selling author whose Landmark books (00:00:49) on the history and future of humanity (00:00:51) have sold an astonishing 45 million (00:00:54) copies and made him the public (00:00:56) intellectual of our time this is the (00:00:58) first time that we are basically about (00:01:00) to enter a nonhuman culture the big (00:01:04) question is whether we will force it to (00:01:06) slow down or it will force us to speed (00:01:10) up until the moment we collapse and die (00:01:13) his latest book and the terrain for (00:01:15) today's conversation is Nexus an (00:01:18) absolutely essential read that makes (00:01:21) quite a compelling case for why (00:01:22) artificial intelligence will be the (00:01:24) biggest disruption in the history of (00:01:27) civilization AI can make decisions they (00:01:30) they are not just Tools in our hands (00:01:32) they are agents creating new realities (00:01:35) it's very difficult to appreciate the (00:01:36) dangers because the dangers they're kind (00:01:39) of alien in the Hollywood scenario you (00:01:41) have the Killer Robots shooting people (00:01:44) in real life it's the humans pulling the (00:01:46) trigger but the AI is choosing the (00:01:55) [Music] (00:01:57) targets thank you for coming I (00:01:59) appreciate you being here today I'm (00:02:01) excited to unpack what I think is a a (00:02:04) really uh revelatory book a very (00:02:06) important book that speaks to perhaps (00:02:08) the most vital issue of our time and in (00:02:11) reflecting upon it I was thinking back (00:02:13) on homod deas which came out in (00:02:16) 2015 yeah and in that book you address (00:02:19) AI uh but at that time it was as if you (00:02:23) were sounding an alarm on a future story (00:02:26) uh that had yet to be had yet to be (00:02:29) written yeah and perhaps it came off a (00:02:31) bit Cassandra you know in that moment (00:02:34) and I'm curious as we find ourselves now (00:02:36) in 2024 eight nine years later it's as (00:02:40) if not only are we you know kind of on (00:02:43) the cusp of this new Revolution we're (00:02:45) mired in it in a way that perhaps even (00:02:48) is far more intense than even you (00:02:51) predicted at that time yeah I mean (00:02:53) things have been moving much much faster (00:02:55) than I think any of us (00:02:57) predicted and you know in 2016 AI was (00:03:01) like this tiny Cloud on the horizon that (00:03:05) might arrive in decades or even (00:03:07) centuries and here we are in 2024 and (00:03:10) the storm is kind of upon us and I think (00:03:13) maybe the most important thing is is (00:03:15) really to understand what what AI is (00:03:17) because now there is so much hype around (00:03:21) AI that it's becoming difficult for (00:03:23) people to understand what is AI now (00:03:26) everything is AI you know especially in (00:03:28) the in the markets in in the investment (00:03:31) World they attach the tag AI to just (00:03:34) about anything in order to sell it so (00:03:37) you know your coffee machine is now a (00:03:39) coffee machine is an AI coffee machine (00:03:41) and your shoes are AI shoes and what is (00:03:45) AI so you know the key thing to (00:03:48) understand is that AIS are able to learn (00:03:52) and change by themselves to make (00:03:55) decisions by themselves to invent new (00:03:58) ideas by themselves if a machine cannot (00:04:00) do that it's not really an AI so a (00:04:03) coffee machine that just makes you (00:04:05) coffee automatically but by a (00:04:08) pre-programmed way and it never learns (00:04:10) anything new it's just an automatic (00:04:12) machine it's not an AI it becomes an AI (00:04:15) if as you approach the coffee machine (00:04:18) the machine before you press any button (00:04:21) addresses you and says to you I've been (00:04:23) watching you for the last weeks or (00:04:26) months and based on everything I've (00:04:29) learned about you (00:04:30) and your facial expression and the time (00:04:32) of day and so forth I predict you would (00:04:35) like an espresso so I already took the (00:04:38) liberty to make a cup for you he made (00:04:40) the decision independently and it's (00:04:42) really an AI if it then tells you (00:04:45) actually I've invented a new machine a (00:04:48) new beverage a new drink that no human (00:04:52) ever thought about before I call it (00:04:54) bestpresso and I think it's better than (00:04:57) espresso you would like it more and I (00:04:59) took the Liber to prepare a cup for you (00:05:01) then it's really an AI something that (00:05:04) can make decisions and invent new ideas (00:05:06) by itself and therefore by definition (00:05:09) something that we cannot predict how it (00:05:12) will develop and evolve and for good or (00:05:15) or for bad it can invent medicines and (00:05:18) treatments we never thought about but it (00:05:20) can also invent weapons and dangerous (00:05:24) strategies that go beyond our (00:05:27) imagination you characterize AI not as (00:05:30) artificial intelligence but as alien (00:05:33) intelligence you give it a different (00:05:34) term can you explain the difference (00:05:37) there and why you why you've landed on (00:05:39) that word yeah traditionally the acronym (00:05:43) AI stood for artificial intelligence but (00:05:46) with every passing year AI becomes less (00:05:49) artificial and more alien alien not in (00:05:53) the sense that it's coming from out of (00:05:54) space it's not uh we create it but alien (00:05:57) in the sense it analyzes information (00:06:00) makes decisions invents new things in a (00:06:03) fundamentally different way than human (00:06:05) beings MH again artificial is from (00:06:08) artifact it give us the impression that (00:06:11) this is an artifact that we control and (00:06:14) this is misleading because yes we (00:06:16) designed the kind of baby AI we we gave (00:06:20) them the ability to learn and change by (00:06:22) themselves and then we released them to (00:06:24) the world and they do things that are (00:06:28) not under our control that are (00:06:30) unpredictable and in this sense they are (00:06:33) alien and again I mean humans are (00:06:36) organic entities like other animals we (00:06:39) function organically for instance we (00:06:42) function by Cycles day and night summer (00:06:45) and winter we sometimes active sometimes (00:06:48) we need to rest we need to sleep AIS are (00:06:52) alien in the sense that they are not (00:06:54) organic they function in a completely (00:06:57) different way not by cycles and they (00:06:59) don't need to rest and they don't need (00:07:01) to sleep and now as they take over more (00:07:04) and more parts of reality parts of (00:07:06) society there is a kind of tgof War of (00:07:10) who would be forced to adapt to whom (00:07:13) would the inorganic AIS be forced to (00:07:16) adapt to the organic cycles of the human (00:07:19) body of the human being or would humans (00:07:22) be pressured into adopting this kind of (00:07:25) inorganic lifestyle and starting with (00:07:28) the simplest thing that a I are always (00:07:30) on but people need time to be off so if (00:07:34) you think even about something like the (00:07:36) financial markets traditionally if you (00:07:38) look at Wall Street it's open only (00:07:41) Mondays to Fridays 9:30 in the morning (00:07:44) to 4:00 in the afternoon it's off for (00:07:46) the night it's off for the weekends it (00:07:49) takes vacations on Christmas on (00:07:51) Independence Day and now as algorithms (00:07:55) and AIS are taking over the markets (00:07:58) they're always on and this puts pressure (00:08:00) on human bankers and Investments and so (00:08:03) forth you can't take a minute off (00:08:05) because then you're left behind so in (00:08:08) this sense they are alien not in the (00:08:09) sense that they came for Mars to (00:08:12) understand artificial intelligence and (00:08:14) to understand what is actually happening (00:08:17) and where we're heading the thesis of of (00:08:20) this latest book requires us to (00:08:23) understand the nature of information (00:08:25) itself and the formative ways in which (00:08:26) the evolution of information networks (00:08:29) are inext (00:08:30) from the evolution and progress of (00:08:33) humankind so I'm curious about how you (00:08:36) discovered that lens into kind of (00:08:38) understanding the nature of artificial (00:08:40) intelligence and why it's important to (00:08:43) contextualize what is occurring right (00:08:45) now through that (00:08:47) perspective it's actually something I (00:08:49) began exploring in in previous books the (00:08:52) ideas is that uh information is the most (00:08:56) fundamental stratum most fundamental (00:08:59) basis of human society and of human (00:09:01) reality cuz the human superpower is the (00:09:04) ability to cooperate in very large (00:09:06) numbers if you compare us to chimpanzees (00:09:09) to elephants to hyenas individually (00:09:12) there are some things I can do in the (00:09:14) chimpanze con and vice versa uh our big (00:09:17) Advantage is not on the individual level (00:09:20) the really big Advantage is that (00:09:22) chimpanzees can cooperate in you know a (00:09:24) few dozen chimpanzees like 50 (00:09:26) chimpanzees can cooporate maybe a 100 (00:09:29) but with humans with Homo sapiens there (00:09:31) is no limit we can cooperate in (00:09:34) thousands in millions in billions if you (00:09:36) think about the World Trade Network like (00:09:39) the food we eat the shoes we wear (00:09:42) everything we consume it sometimes come (00:09:44) from the other side of the world so if (00:09:45) you have 8 billion people cooperating (00:09:49) and this is our big advantage over the (00:09:52) chimpanzees and all the other animals (00:09:54) what makes it possible for us to (00:09:56) cooperate with millions and billions of (00:09:59) other human beings it's information (00:10:01) information is what holds all these (00:10:04) large scale systems together and to (00:10:07) understand human history is to a large (00:10:09) extent to understand the flow of (00:10:12) information and I'll give an example if (00:10:14) you think for instance about the (00:10:16) difference between democracies and (00:10:18) dictatorships we tend to think about it (00:10:21) as a difference or as a conflict between (00:10:23) values between ethical (00:10:26) systems democracies believe in Freedom (00:10:29) dictatorships believe in hierarchies (00:10:31) things like that and which is true as (00:10:33) far as it goes but on a deeper level (00:10:36) information flows differently in (00:10:39) democracies and dictatorships it's a (00:10:41) different shape a different kind of an (00:10:44) Information Network in a (00:10:46) dictatorship all decisions are made (00:10:49) centrally dictatorships come from (00:10:51) dictate one person dictates everything (00:10:54) Putin dictates everything in Russia Kim (00:10:56) junun dictates everything in North Korea (00:10:58) so all the information flows to a single (00:11:02) Hub where all the decisions are being (00:11:04) made and sent back as orders so it's a (00:11:07) very centralized Information Network a (00:11:10) democracy on the other hand if you look (00:11:13) at it in terms of you're in alter space (00:11:15) looking at the flow of information in (00:11:17) the United States you will see several (00:11:20) centers in the country Washington the (00:11:24) political Center New York the Financial (00:11:26) Center Los Angeles the maybe autistic (00:11:29) Center (00:11:30) but there is no single Center that (00:11:32) dictates everything you have several (00:11:34) centers and you also have lots and lots (00:11:37) of smaller hubs and centers where (00:11:40) decisions are constantly being made (00:11:43) private corporations private businesses (00:11:46) voluntary associations individuals (00:11:49) making lots of decisions constantly (00:11:51) exchanging information without that (00:11:54) information ever having to pass through (00:11:57) the center through Washington or even (00:11:59) even through New York or even through (00:12:01) Los Angeles so just looking you don't (00:12:04) know anything about the values of the (00:12:05) people you just imagine you're in outer (00:12:08) space on in some spaceship or satellite (00:12:10) just observing the flow of information (00:12:13) down below the planet you will see that (00:12:16) North Korea is very different (00:12:18) information flow than the United States (00:12:23) and this is crucial to understand and (00:12:25) when you look at thousands of years of (00:12:27) history and how history changes and (00:12:30) different regimes rise and (00:12:32) fall understanding what kind of (00:12:34) information technology is available is a (00:12:37) key to understanding which political (00:12:40) systems or economic systems win for most (00:12:43) of History a large scale democracy like (00:12:46) the United States was simply (00:12:49) impossible if you think about the (00:12:50) ancient world the only examples we know (00:12:53) of democracy are small city states like (00:12:56) Republican Rome or like ancient Athens (00:12:59) or even smaller tribes we don't have any (00:13:02) example of a large scale democracy of (00:13:05) millions of people spread over a vast (00:13:08) territory that function democratically (00:13:11) now we know the stories for instance (00:13:13) about the fall of the Roman Republic and (00:13:15) the rise of the Caesars of the Emperors (00:13:18) of the autocrats but it's really not the (00:13:21) fault of Augustus Caesar or Nero or any (00:13:24) of the other Emperors that Rome became (00:13:27) an autocratic Empire simply there was no (00:13:31) way that the information technology (00:13:33) necessary to maintain a large scale (00:13:36) democracy which is bigger than just the (00:13:38) city of Rome like the all of Italy or (00:13:41) the all of the Mediterranean democracy (00:13:43) is a conversation and how can millions (00:13:46) of people spread over thousands of (00:13:48) kilometers Converse and decide whether (00:13:52) to go to war with the Persian Empire (00:13:54) what to do about the immigration crisis (00:13:56) on the danu with all these Germans (00:13:58) trying to get in you can't have a (00:14:00) conversation because you don't have the (00:14:02) information technology and you know if (00:14:05) it was just the fault of Caesar that (00:14:07) Rome became an autocratic Empire we (00:14:09) should have seen some other examples of (00:14:12) a large scale democracy in India in (00:14:15) China somewhere but nowhere we only (00:14:18) begin to see large scale democracies in (00:14:20) the late modern era after the rise of (00:14:24) new information Technologies which were (00:14:26) not available to the Romans like the (00:14:28) printed newspaper (00:14:30) and then the Telegraph and the radio and (00:14:32) television and so forth once you have (00:14:35) these Technologies you begin to see (00:14:36) large scale democracies like the United (00:14:38) States and one final Point why is it so (00:14:41) important to understand this once you (00:14:44) understand that democracy is actually (00:14:46) built on top of Information Technology (00:14:49) you also begin to understand the current (00:14:51) crisis of democracy because you know now (00:14:54) all over the world not just in the US we (00:14:56) have a crisis of democracy and to to a (00:14:59) large extent this is because there is a (00:15:01) new information technology social media (00:15:04) algorithms AIS and it's like you know (00:15:08) you're changing the basis of everything (00:15:10) so there it's no wonder there is an (00:15:12) earthquake in the structure that is (00:15:14) built on top of it so we have this idea (00:15:17) that the Advent or the Improvement of (00:15:21) information systems and information (00:15:23) technology is part and parcel of the (00:15:27) empowerment of democratic systems across (00:15:29) the world but built into that is this (00:15:32) sort of indelible misconstrual of (00:15:35) information this assumption or (00:15:38) presumption that more information is (00:15:40) better and leads to truth and knowledge (00:15:44) and wisdom uh and your book kind of puts (00:15:48) the lie to that and tells a very (00:15:50) different story around not only the (00:15:52) definition of information but its (00:15:54) purpose yeah I mean information isn't (00:15:57) truth information is connection it's (00:16:00) something that holds a lot of people (00:16:03) together and unfortunately what we see (00:16:06) in history that it's often much easier (00:16:09) to connect people to create social order (00:16:13) with the help of Fiction and Fantasy and (00:16:16) propaganda and lies than with the truth (00:16:20) so most information is not (00:16:22) true uh the truth is a very rare subset (00:16:27) of the information in the world (00:16:30) the problem of Truth is that the truth (00:16:32) first of all is costly whereas fiction (00:16:34) is very cheap if you want to write a (00:16:37) truthful history book about the Roman (00:16:39) Empire for instance you need to invest a (00:16:42) lot a lot of energy time money you need (00:16:45) to study Latin you probably need to (00:16:47) study Greek ancient Greek you need to do (00:16:50) archaeological excavations and find (00:16:52) these ancient whether inscriptions or (00:16:55) Pottery or weapons and analyze them very (00:16:59) cost ly and difficult to write a (00:17:01) fictional story about the Roman Empire (00:17:03) very easy you just write anything you (00:17:04) want and it's there on on the on the (00:17:06) page or on the Internet the truth is (00:17:09) often also very complicated because (00:17:11) reality is complicated you want to give (00:17:14) a truthful explanation for why the Roman (00:17:16) Republic fell or why the Roman Empire (00:17:19) eventually fell very complicated whereas (00:17:22) fiction can be made as easy as as simple (00:17:24) as possible and people tend to prefer (00:17:26) simple explanations over complicated (00:17:29) ones and finally the truth can be (00:17:33) painful (00:17:34) unattractive we often don't want to know (00:17:36) the truth about ourselves whether as (00:17:38) individuals which is why we go to (00:17:40) therapy for many years to know the (00:17:42) things we don't want to know about (00:17:44) ourselves and also on the level of (00:17:46) entire nations you know each nation has (00:17:48) its own Dark episodes its own skeletons (00:17:52) or cemeteries in the closet that people (00:17:54) don't want to know about a politician (00:17:57) that you know in an election campaign (00:17:58) would just tell tell people the truth (00:18:00) the whole truth and nothing but the (00:18:02) truth is unlikely to win many (00:18:04) votes uh so in this (00:18:07) competition between the truth which is (00:18:10) costly and complicated and sometimes (00:18:12) painful and fiction which is cheap and (00:18:17) simple and you can make it very (00:18:18) attractive fiction tends to win and if (00:18:21) you look at you know the the the large (00:18:24) scale systems networks in history (00:18:27) they're often built on fictions not on (00:18:31) the truth maybe I I give one example if (00:18:34) you think about visual information like (00:18:37) portraits paintings (00:18:39) photographs um so what is the most (00:18:42) common portrait in the world what is the (00:18:44) most famous face in the history of (00:18:47) humanity it is the face of Jesus I mean (00:18:50) there are more portraits of Jesus than (00:18:53) of any other person in the history of (00:18:55) the world billions and billions produced (00:18:57) over centuries in Cath (00:18:59) and churches and homes and fully 100% of (00:19:04) them are fictional there is not a single (00:19:07) authentic truthful portrait of Jesus (00:19:10) anywhere uh we have no portrait of him (00:19:13) from his own (00:19:14) lifetime uh the Bible doesn't say a (00:19:17) single word about how he looked like (00:19:19) there is not a single word in the Bible (00:19:21) whether Jesus was tall or short uh dark (00:19:24) hair or blonde or bold nothing all the (00:19:28) images and you know it's one of the most (00:19:30) famous faces in history it all comes (00:19:32) from the human (00:19:33) imagination and it's still very (00:19:36) successful in inspiring people and (00:19:38) uniting people could be for good (00:19:40) purposes you know charity and building (00:19:43) hospitals and helping the poor but could (00:19:45) also be for bad purposes Crusades (00:19:48) persecutions inquisitions but either way (00:19:52) the the immense power of of a fictional (00:19:55) image to unite people and going looking (00:19:59) what's happening today in the world so (00:20:01) you have these you know big tech (00:20:03) companies and social media companies (00:20:05) that they tell us that all information (00:20:07) is always good so let's remove all (00:20:10) restrictions on the flow of information (00:20:12) and flood the world with more and more (00:20:14) information and more information would (00:20:16) mean more truth more knowledge more more (00:20:18) wisdom and this is simply not true most (00:20:22) information is actually junk if you just (00:20:24) flood the world with information the (00:20:27) truth will sink to the bottom it will (00:20:29) not rise to the top again because it's (00:20:31) costly and (00:20:33) complicated and you look around we have (00:20:36) this flood of information we have the (00:20:39) most sophisticated information (00:20:41) technology in history and people are (00:20:43) losing the ability to hold a (00:20:45) conversation to talk and listen to one (00:20:48) another you know in the United States (00:20:50) Republicans and Democrats are barely (00:20:52) able to to talk to each other and it's (00:20:54) not an American phenomena you see the (00:20:56) same thing in in Brazil in France in in (00:20:59) the Philippines all over the world (00:21:01) because again the basic misconception is (00:21:04) that more information is always good for (00:21:05) us it's like thinking that more food is (00:21:08) always good for us and most information (00:21:11) is junk information yeah and what's (00:21:14) Curious to me about all of this is that (00:21:16) on some level what you're saying is (00:21:18) there's nothing new about this there is (00:21:20) this idea that suddenly we found (00:21:22) ourselves in a post-truth world and part (00:21:25) of what you're saying is it's kind of (00:21:27) always been that way but the qualitative (00:21:29) difference right now is not by (00:21:31) definition these platforms that allow us (00:21:34) to share information as much as it is (00:21:37) the algorithms that Empower them that (00:21:39) make the decisions about what we're (00:21:41) seeing and when we're seeing it yeah I (00:21:43) mean this is maybe the first place you (00:21:46) see the power of AIS to make independent (00:21:51) decisions in a way that reshapes the (00:21:53) world when I said earlier that you know (00:21:56) AI can make decisions and AI they are (00:21:59) not just Tools in our hands they are (00:22:01) agents creating new realities so you may (00:22:04) think okay this is a prophecy for the (00:22:06) future a prediction about the future but (00:22:09) it's already in the past because even (00:22:12) though social media algorithms they are (00:22:14) very very primitive AIS you know the (00:22:17) fair generation of AIS they still (00:22:20) reshaped the world with the decisions (00:22:22) they made in social media on Facebook (00:22:25) Twitter Tik Tok all that the ones that (00:22:29) make the decision what you will see at (00:22:31) the top of your news feed or the next (00:22:35) video that you'll be recommended It's (00:22:37) Not a Human Being sitting there making (00:22:39) these decisions it's an AI it's an (00:22:42) algorithm and these algorithms were (00:22:45) given a relatively simple and seemingly (00:22:48) benign goal by the (00:22:50) corporations the goal was increase user (00:22:54) engagement which means in simple English (00:22:56) make people spend more time on the (00:22:58) platform (00:22:59) uh because the more time people spend on (00:23:01) Tik Tok or Facebook or Twitter or (00:23:03) whatever the company makes more money it (00:23:05) sells more advertisements it harvests (00:23:08) more data that it can then sell to third (00:23:10) parties so more time on the platform (00:23:12) good for the company this is the goal of (00:23:14) the algorithm now engagement sounds like (00:23:18) a good thing who doesn't want to be (00:23:20) engaged but the algorithms then (00:23:23) experimented on billions of human guinea (00:23:26) pigs and discovered something which (00:23:29) which was of course discovered even (00:23:30) earlier by humans but now the algorithms (00:23:32) discovered it the algorithms discovered (00:23:35) that the easiest way to increase user (00:23:38) engagement the easiest way to grab (00:23:40) people's attention and keep them glued (00:23:42) to the screen is by pressing the greed (00:23:46) or hate or fear button in our minds you (00:23:51) show us some hate filled conspiracy (00:23:53) theory and we become very angry we want (00:23:55) to to see more we tell about it to all (00:23:57) our friends us their engagement goes up (00:24:01) and this is what they did over the last (00:24:03) 10 or 15 years they flooded the world (00:24:06) with hate and greed and fear which is (00:24:09) why again the conversation is breaking (00:24:12) down very hard to hold a conversation (00:24:14) with all this hate and fear yeah it's a (00:24:17) function of unintended consequences that (00:24:20) on some level is no different than Nick (00:24:22) bostrom's you know alignment problem you (00:24:24) know thought experiment about paper (00:24:26) clips like this is the exact same thing (00:24:29) and I think it speaks to not only human (00:24:32) ignorance but human hubris around this (00:24:34) powerful technology I think you know you (00:24:37) talk so much about stories and how (00:24:38) indelible they are in terms of crafting (00:24:41) our reality but one of those stories is (00:24:44) we know what we're doing we can handle (00:24:46) it we understand the consequences we (00:24:49) know the downside here and we're making (00:24:51) sure that what we're putting out into (00:24:53) the world is is safe and consumer (00:24:55) friendly when you know on some level (00:24:57) they know it's not but Al they have no (00:25:00) idea you know what will become of it as (00:25:03) a result and so we're just in this (00:25:05) Frontier this unregulated Frontier where (00:25:09) anything goes at the moment yeah I mean (00:25:11) I think it's important what you said (00:25:14) that these are kind of unintended (00:25:16) consequences like the people who manage (00:25:18) the social media companies they are not (00:25:20) evil they didn't set out to destroy (00:25:23) democracy or to flood the world with (00:25:25) with hate and and and so forth um they (00:25:28) just really didn't foresee that when (00:25:30) they give the algorithm the goal of (00:25:33) increasing user engagement the algorithm (00:25:35) will start to promote hate and one of (00:25:38) the first places that let me just (00:25:40) interject quickly on that though now (00:25:42) that they know that that's the case it's (00:25:44) not as if they're backtracking that's (00:25:46) true they're EXA they're not exactly (00:25:48) regulation friendly at the moment no (00:25:50) absolutely not so all right sorry go (00:25:52) ahead you're right now they know and (00:25:54) they are not doing nearly enough but (00:25:57) initially when they started the whole (00:25:59) ball rolling they really didn't know and (00:26:02) one of the places you saw it for the (00:26:03) first time this was you know eight years (00:26:06) ago when I published homo this was (00:26:08) happening I I didn't pay attention to it (00:26:10) either in Myanmar bur Burma the country (00:26:13) formerly known as Burma Facebook was (00:26:16) basically the internet and and cly the (00:26:19) biggest social media uh platform and uh (00:26:23) in the 2010s the algorithms of Facebook (00:26:26) in Myanmar they deliberately spread (00:26:30) terrible conspiracy theories and fake (00:26:32) news about the rohinga minority in (00:26:35) Myanmar which led to an ethnic with of (00:26:38) course it was not the only reason there (00:26:40) was deep-seated hatred towards rohinga (00:26:42) much before but this kind of propaganda (00:26:45) campaign online on Facebook contributed (00:26:48) to an ethnic cleansing campaign between (00:26:51) 2016 and 2017 2018 in which thousands of (00:26:55) rohinga were killed tens of thousands (00:26:58) were raped and hundreds of thousands (00:27:01) were expelled you now have close to a (00:27:03) million rohinga refugees in in (00:27:05) Bangladesh and elsewhere and this was (00:27:08) fueled to a large extent by this (00:27:10) conspiracy theories and fake news on (00:27:12) Facebook and at the time the executive (00:27:15) of Facebook had no I mean they didn't (00:27:18) know even the rohinga existed it's not (00:27:21) like it was a conspiracy of Facebook (00:27:23) against them for the Hall of (00:27:25) Myanmar a country where Facebook had (00:27:28) Millions and millions of (00:27:30) users they by 2018 this is after they (00:27:34) got reports of the of the ethnic (00:27:36) cleansing campaign they had just a (00:27:38) handful of humans trying to kind of (00:27:43) regulate uh the actions of millions of (00:27:46) users in the (00:27:48) algorithms and they didn't even speak (00:27:50) boures like when the algorithm chose (00:27:53) okay I I'll show people this hatefi (00:27:56) conspiracy theory video in buor (00:27:59) nobody in Facebook headquarters spoke (00:28:01) bmes they had no idea what the algorithm (00:28:05) was promoting the key thing is is not to (00:28:08) absolve the humans from responsibility (00:28:11) it's to understand that even very (00:28:14) primitive AIS and we were talking about (00:28:16) you know like eight years ago MH not (00:28:19) things like CHP to still the the (00:28:22) decisions made by these algorithms to (00:28:24) promote certain content had far reaching (00:28:27) and terrible consequen quences in (00:28:29) Myanmar they were not just producing (00:28:31) conspiracy theories they were producing (00:28:33) their millions of users producing you (00:28:35) know cooking lessons and biology lessons (00:28:38) and sermons on compassion from Buddhist (00:28:41) monks and conspiracy theories and the (00:28:43) algorithms made a decision to promote (00:28:46) the conspiracy theories and this is just (00:28:49) kind of a warning of look what happens (00:28:52) with even very primitive AIS and the AIS (00:28:56) of today which are far more (00:28:57) sophisticated than (00:28:59) 2016 they too are still just the very (00:29:02) early stages of the AI evolutionary (00:29:05) process and we can think about it like (00:29:07) the evolution of of animals until you (00:29:10) get to humans you have 4 billion years (00:29:14) of evolution you start with (00:29:15) microorganisms like amibas and it took (00:29:18) billions of years of evolution to get to (00:29:22) dinosaurs and mammals and humans now AIS (00:29:25) are present at the beginning of a (00:29:27) parallel process (00:29:29) the CH GPT and so forth they are the (00:29:32) amibas of the AI world but AI evolution (00:29:35) is not organic it's inorganic it's (00:29:38) digital and it's millions of times (00:29:41) faster so where it took billions of (00:29:43) years to get from amibas to dinosaurs it (00:29:46) might take just 10 or 20 years to get (00:29:49) from the AI amibas of today to AI T-Rex (00:29:54) in 2040 or 2050 maybe even less maybe (00:29:57) even less we're talking about I don't (00:29:58) think our brains are are (00:30:01) organized properly to really comprehend (00:30:04) The Accelerated speed at which this is (00:30:07) self-learning and iterating and (00:30:09) improving upon itself like just it's a (00:30:11) compounding thing that is astronomical (00:30:14) meanwhile trillions of dollars are being (00:30:16) spent to build these server Farms with (00:30:18) these Nvidia chips and there's so much (00:30:20) power required to keep these things (00:30:22) going they're talking about nuclear I (00:30:24) mean this is like this is a whole new (00:30:27) world and yet in talking about it it (00:30:30) still feels somewhat like an academic (00:30:33) exercise because for myself or somebody (00:30:36) who might be watching or listening their (00:30:38) experience with AI comes in the form of (00:30:41) chat GPT or some of these helpful tools (00:30:44) like I like my algorithm it shows me the (00:30:47) kind of products that I want to buy (00:30:49) without having to search for it and a (00:30:51) simple example would be preparing for (00:30:54) this podcast like I listen to your book (00:30:55) on audiobook and I'm doing what I (00:30:57) usually do pulling up a bunch of tabs (00:30:59) and you know like just collating a bunch (00:31:01) of information on you and the book and (00:31:03) the message that you're putting out but (00:31:05) I did something I had never done before (00:31:07) which is I got a PDF of Nexus and I (00:31:10) uploaded it to a tool called notebook LM (00:31:13) M and that tool then synopsized the (00:31:16) entire book and created a chat bot where (00:31:19) I could ask it questions about your book (00:31:21) and ask it to elaborate on certain (00:31:23) Concepts and it will even create a (00:31:26) podcast conversation between two people (00:31:29) about the subject matter of the (00:31:31) book so even this conversation is at (00:31:34) risk right irrelevant and I'm like wow (00:31:36) that's kind of a a remarkably helpful (00:31:38) tool and it's easy to to you know just (00:31:42) not really appreciate or connect with (00:31:45) the downside risk and power of these (00:31:48) tools and where they're leading us so I (00:31:50) think what I'm saying is I guess the (00:31:52) point I'm trying to make is consumers (00:31:54) like all of us we're we're being lured (00:31:57) into a Trust of something so powerful we (00:32:01) can't comprehend and are ill equipped to (00:32:04) be able to kind of cast our gaze into (00:32:05) the future and imagine where this is (00:32:08) leading us absolutely I mean part of it (00:32:10) is that there is enormous positive (00:32:12) potential in AI it's not like it's all (00:32:15) doom and gloom there is really enormous (00:32:17) positive potential if you think about (00:32:18) the implications for healthc care that (00:32:20) you know AI doctors available 24 hours a (00:32:23) day that know our entire medical history (00:32:27) and have read every medical paper that (00:32:29) was ever published and can tailor their (00:32:32) advice their treatment to our specific (00:32:35) life history and our blood pressure our (00:32:38) genetics it it can be the biggest (00:32:40) revolution in healthcare ever if you (00:32:42) think about self-driving Vehicles so (00:32:44) every year more than a million people (00:32:47) die all over the world in car accidents (00:32:49) most of them are caused by human error (00:32:52) like people drinking and then driving or (00:32:54) falling asleep at the wheel or whatever (00:32:56) uh sell driving vehicles are likely to (00:32:58) sell save about a million lives every (00:33:00) year this is amazing you think about (00:33:02) climate change so yes developing the AIS (00:33:04) will consume a lot of energy but they (00:33:07) could also find new sources of energy (00:33:09) new ways to to harness energy that could (00:33:12) be our best shot at at preventing (00:33:14) ecological collapse uh so there is (00:33:17) enormous positive potential we shouldn't (00:33:19) deny that we should be aware of it and (00:33:21) on the other hand it's very difficult to (00:33:23) appreciate the dangers because the (00:33:25) dangers again they are kind of alien (00:33:27) like if you think about nuclear energy (00:33:29) yeah also had positive potential nuclear (00:33:31) cheap nuclear energy but people had a (00:33:34) very good grasp of the danger nuclear (00:33:36) war anybody can understand the danger of (00:33:38) that with AI it's much more complex (00:33:42) because the danger is not (00:33:43) straightforward the danger is really I (00:33:46) mean we we've seen the Hollywood science (00:33:48) fiction scenarios of the big robot (00:33:50) Rebellion that one day the big computer (00:33:53) or the AI decides to take over the world (00:33:56) and kill us or enslave us (00:33:59) and this is extremely unlikely to happen (00:34:00) anytime soon because the AIS are still a (00:34:03) kind of very narrow intelligence like (00:34:06) the AI that can summarize a book it it (00:34:09) doesn't know how to act in the physical (00:34:11) world outside you have AIS that can fold (00:34:13) proteins you have ai that can play chess (00:34:16) but we don't have this kind of General (00:34:18) AI that can just find its way around the (00:34:20) world and build the robot army and and (00:34:23) whatever so people it it's how to (00:34:26) understand so what's so dangerous about (00:34:28) something which is so kind of narrow in (00:34:30) its abilities and I would say that the (00:34:33) danger doesn't come from the big robot (00:34:35) Rebellion it comes from the AI (00:34:37) bureaucracies already today and more and (00:34:40) more we will have not one big AI trying (00:34:42) to take over the world we will have (00:34:44) millions and billions of AIS constantly (00:34:47) making decisions about us everywhere you (00:34:50) apply to a bank to get a loan it's an AI (00:34:53) deciding whether to give you a loan you (00:34:55) apply to get a job it's an AI deciding (00:34:57) whether to give you a job you're in (00:34:59) court you're found guilty of some crime (00:35:01) the AI will decide whether you go for 6 (00:35:04) months or 3 years or whatever even in (00:35:07) armies we already see now in the war in (00:35:09) Gaza in with the war in Ukraine AI make (00:35:11) the decision about what to bomb um and (00:35:14) in the Hollywood scenario you have the (00:35:17) Killer Robots shooting people in real (00:35:19) life it's the humans pulling the trigger (00:35:22) but the AI is choosing the targets is (00:35:25) telling them what to this is much more (00:35:27) complex yeah then the standard (00:35:35) scenario every point of connection with (00:35:38) bureaucracy then becomes turned over to (00:35:41) an algorithm that makes decisions in a (00:35:43) black box without the opportunity for (00:35:47) rebuttal or conversation right so we (00:35:49) we're Outsourcing all of these decisions (00:35:51) and creating like an autocratic diaspora (00:35:54) of decision makers right and that in (00:35:56) turn like you can imagine over time like (00:35:59) what emerges from that is is like a (00:36:02) godhead or a Pantheon of gods where (00:36:05) there's an authoritarian regime that's (00:36:07) dispersed across this in which we are (00:36:10) relenting our agency over to these (00:36:12) machines and trusting that they're (00:36:15) making the right decisions but not (00:36:17) knowing how those decisions are being (00:36:18) made even the engineers who are creating (00:36:20) the algorithms don't know and there's (00:36:22) something you know kind of innately (00:36:24) terrifying about that again it's not (00:36:26) authoritarian in the sense that there is (00:36:28) a single human being that is kind (00:36:30) pulling all the levers no it's it's the (00:36:32) AI like the bank has this AI that (00:36:34) decides who is qualified to get a loan (00:36:37) and if they tell you we decided not to (00:36:39) give to give you a loan and you ask the (00:36:41) bank why not and the bank says we don't (00:36:43) know I mean computer says no I mean the (00:36:45) algorithm says no we don't understand (00:36:47) why the algorithm says no but we trust (00:36:50) the algorithm and this is likely to (00:36:53) spread to to more and more places the (00:36:56) key thing is it's not that the bank is (00:36:58) hiding something from you it's really (00:37:01) that the AIS make decisions in a very (00:37:03) different way than human beings on a (00:37:06) basis of a lot more data so if the bank (00:37:10) really wanted to explain to you why they (00:37:12) refused to give you a loan like let's (00:37:14) say there is a law the government passes (00:37:16) a law of a right to an explanation if (00:37:19) the bank refused to give you a loan you (00:37:21) can apply they must give you an (00:37:23) explanation so the explanation well (00:37:25) people fear that it will be kind of I (00:37:27) don't know racist bias or homophobic (00:37:30) bias like in the old days that the (00:37:32) algorithm so that you're black or you're (00:37:34) Jewish or you're gay and this is why I (00:37:36) refuse to give you a loan it won't be (00:37:38) like that I mean the bank will send you (00:37:41) an entire encyclopedia in millions of (00:37:43) pages saying this is why the computer (00:37:46) refused to give you a loan the computer (00:37:48) took into account thousands and (00:37:51) thousands of data points about you each (00:37:54) one based on statistics on millions of (00:37:57) of PR previous cases and now you can go (00:38:00) over these millions of pages if you like (00:38:03) and if you want to challenge okay but (00:38:05) but it's not the kind of old style (00:38:08) racism or whatever sure a new version of (00:38:12) the terms and conditions that we just (00:38:14) click on without reading right except uh (00:38:16) extrapolated hundredfold um in addition (00:38:19) to that with all of these data points I (00:38:22) can't help but think that that you know (00:38:24) these these (00:38:25) machines the veracity of the information (00:38:28) that these machines provide us with is (00:38:31) only as reliable as the data sets that (00:38:35) it has been provided with and and right (00:38:38) now we're tipto into a situation where (00:38:42) the internet is being uh rapidly (00:38:45) degraded because it's being populated (00:38:48) more and more by AI content now when you (00:38:51) go to Google and you search the first (00:38:53) thing you see is a is sort of an AI kind (00:38:55) of summary of your query as opposed to (00:38:59) links and this in turn is undermining (00:39:02) the business model of Legacy Media and (00:39:05) all forms of media right so as those (00:39:07) continue to die on the vine more and (00:39:10) more of the internet will be a result of (00:39:12) AI generated content and then it becomes (00:39:14) a recursive thing in which it's feeding (00:39:17) upon its own inputs to make decisions (00:39:20) and you know with that like you can (00:39:23) imagine a degradation of the data set (00:39:26) upon which it is making those decisions (00:39:29) exactly even if you think about (00:39:30) something like music so AI that now (00:39:33) creates music it basically ate the whole (00:39:36) of human music like for thousands of (00:39:38) years humans produced music or art or (00:39:40) theater whatever within a year the (00:39:43) current AI just ate the whole of it and (00:39:46) digested it and start now creating new (00:39:49) music or new texts or new images and the (00:39:53) first kind of generation of AI texts or (00:39:56) music um this is based on on previous (00:39:59) human culture but with each passing year (00:40:03) the AIS will be eating their own (00:40:06) products because as you know the human (00:40:08) share in music production or the human (00:40:10) share in text production or image (00:40:11) production will go lower and lower most (00:40:15) images most music will be produced at (00:40:17) least to in part by Ai and this will be (00:40:20) the new food that the AI eats and then (00:40:24) you have exactly what you describ this (00:40:25) recursive pattern and where it will lead (00:40:28) us we have no idea I mean another way to (00:40:31) think about it this is the first time (00:40:33) that we are basically about to enter a (00:40:36) non-human (00:40:37) culture like humans are our cultural (00:40:40) entities we live cun inside culture like (00:40:44) all this music and art and also finance (00:40:48) and also religion this is all part of (00:40:50) culture and for tens of thousands of (00:40:52) years the only entities that produced (00:40:54) culture were other humans so all the (00:40:57) songs you ever heard were produced by (00:40:59) humans all the religious mythologies you (00:41:01) ever heard came from the human (00:41:03) imagination now there is a an alien (00:41:06) intelligence a non-human intelligence (00:41:09) that will increasingly produce songs and (00:41:12) music mythology Financial strategies (00:41:15) political (00:41:16) ideas even before we rush to decide is (00:41:19) it good is it bad just stop and think (00:41:22) about the meaning of living in a (00:41:25) nonhuman culture or a culture which is I (00:41:28) don't know 40% or 70% non-human it's not (00:41:32) like going to China and seeing a (00:41:34) different human culture it's like really (00:41:36) alien culture here on Earth yeah my (00:41:38) human mind bristles at that I start (00:41:40) thinking about like this this bias I (00:41:43) have around the originality of human (00:41:46) thought and emotion and this kind of (00:41:48) assumption that AI will never be able to (00:41:51) fully mimic The Human Experience right (00:41:54) there's something indelible about what (00:41:56) it means to be human that the machines (00:41:59) uh will never be able to fully replicate (00:42:01) and when you talk about you know (00:42:04) information the purpose of information (00:42:06) being to create connection a big piece (00:42:10) there is intimacy like intimacy between (00:42:13) human beings so information is meant to (00:42:14) create connection but now we have so (00:42:16) much information and we're feeling very (00:42:19) disconnected so there's something broken (00:42:21) in this system and I think it's driving (00:42:23) this loneliness epidemic but on the (00:42:25) other side it's it's making us value (00:42:29) like intimacy maybe a little bit more (00:42:31) than we were previously uh and so I'm (00:42:34) curious about where intimacy kind of (00:42:36) fits into this you know posthuman World (00:42:39) in which culture is being dictated by (00:42:42) machines I mean human beings are wired (00:42:44) for that kind of intimacy and I think (00:42:45) our radar or our kind of ability to you (00:42:48) know identify it when we see it is part (00:42:51) of what makes us human to begin with (00:42:54) maybe the most important part um I think (00:42:57) the key distinction here that is often (00:42:59) lost is the distinction between (00:43:01) intelligence and (00:43:03) Consciousness that intelligence is the (00:43:05) ability to pursue goals and to overcome (00:43:09) problems and obstacles on the way to the (00:43:11) goal the goal could be a self-driving (00:43:13) vehicle trying to get from here to San (00:43:16) Francisco the goal could be increasing (00:43:18) user user engagement and an intelligent (00:43:22) agent knows how to overcome the problems (00:43:25) on the way to the goal this is (00:43:27) intelligent (00:43:28) and this is something that AI is (00:43:31) definitely acquiring in at least certain (00:43:34) Fields AI is now much more intelligent (00:43:37) than us like in playing chess much more (00:43:41) intelligent than human beings but (00:43:43) Consciousness is a different thing than (00:43:44) intelligence Consciousness is the (00:43:46) ability to feel things pain pleasure (00:43:50) love hate uh when the AI wins a game of (00:43:53) chess it's not joyful if there is a (00:43:56) tense moment in the in the game it's not (00:43:59) clear who is going to win the AI is not (00:44:01) tense it's only the human player which (00:44:03) is tense or frightened or anxious the AI (00:44:07) doesn't feel anything now there is a big (00:44:10) confusion because in humans and also in (00:44:13) other mammals in other animals in dogs (00:44:16) and pigs and horses and whatever (00:44:18) intelligence and Consciousness go (00:44:20) together we solve problems based on our (00:44:23) feelings our feelings are not something (00:44:26) that kind of evolution (00:44:28) decoration it's the core system through (00:44:32) which marals make decisions and solve (00:44:34) problems is based on our feelings so we (00:44:37) tend to think that Consciousness and (00:44:39) intelligence must go together and in all (00:44:41) these science fiction movies you see (00:44:44) that as the computer or robot becomes (00:44:47) more (00:44:48) intelligent then at some point it also (00:44:51) gains Consciousness it falls in love (00:44:53) with the human or (00:44:54) whatever and we have no reason to think (00:44:57) like that yeah Consciousness is not a (00:44:59) mere extrapolation of intelligence a (00:45:02) qualitatively different thing yeah and (00:45:04) again if you think in terms of evolution (00:45:07) so yes the evolution of mammals took a (00:45:10) certain path a certain Road in which you (00:45:14) develop intelligence based on (00:45:17) Consciousness but so far what we see is (00:45:20) computers they took a different (00:45:22) route their Road develops intelligence (00:45:26) without consciousness (00:45:28) I mean computers have been developing (00:45:29) you know for 60 70 years now they are (00:45:31) not very intelligent at least in some (00:45:33) fields and still zero Consciousness now (00:45:37) this could continue indefinitely maybe (00:45:39) they are just on a different path maybe (00:45:41) eventually they will be far more (00:45:43) intelligent than us in everything and (00:45:46) still will have zero Consciousness we'll (00:45:48) not feel pain or pleasure or love or (00:45:51) hate you know the same way that if you (00:45:53) think about birds and (00:45:56) airplanes so airlanes did not become (00:45:58) like birds airlanes don't fly using (00:46:02) feathers and so forth they fly in a (00:46:04) completely different way it's not like (00:46:06) that at a certain point when the (00:46:07) airplane flies fast enough suddenly the (00:46:10) the feathers will appear no and it could (00:46:13) be the same with intelligence and (00:46:14) Consciousness that it will be more and (00:46:17) more intelligent without feelings ever (00:46:20) appearing now what adds to the problem (00:46:23) is that there is nevertheless a very (00:46:26) strong commercial and political (00:46:28) incentive to develop AIS that mimic (00:46:33) feelings to develop AIS that can create (00:46:36) intimate relations with human beings (00:46:39) that can cause human beings to be (00:46:43) emotionally attached to the AIS even if (00:46:46) the AIS have no feelings of themselves (00:46:50) they could be trained they are already (00:46:52) trained to make us feel that they have (00:46:56) feelings mhm and to start developing (00:46:59) relationships with them why is there (00:47:01) such an incentive because intimacy is on (00:47:05) the one hand maybe the most cherished (00:47:08) thing that that the human can (00:47:10) have uh you know I was just on on the (00:47:12) way here we were listening to Barbara (00:47:13) ston singing are people who need people (00:47:17) are the luckiest people in the world (00:47:19) that intimacy is not a liability it's (00:47:22) not something bad that oh I I need this (00:47:24) no it's it's the greatest thing in the (00:47:25) world but it's also potentially the most (00:47:29) powerful weapons weapon in the world if (00:47:32) you want to convince somebody to buy a (00:47:34) product if you want to convince somebody (00:47:36) to vote for a certain politician or (00:47:39) party intimacy is like the Ultimate (00:47:42) Weapon I mean so far in history there (00:47:45) was a big battle for attention how to (00:47:47) grab human attention also we talked (00:47:49) about earlier in social media how how to (00:47:51) get human attention and there were ways (00:47:54) like I don't know in Nazi Germany Hitler (00:47:57) could Force everybody to listen to his (00:47:59) speech on radio so he had command of (00:48:01) attention but not of intimacy there was (00:48:04) no technology for Hitler or Stalin or (00:48:07) anybody else to mass produce intimacy (00:48:11) now is AIS it is possible technically to (00:48:14) mass produce intimacy you can create all (00:48:17) these AIS that will interact with us and (00:48:21) they will understand our feelings (00:48:22) because again feelings are also patterns (00:48:25) You can predict a person's feelings by (00:48:27) watching them for weeks and months and (00:48:29) learning their patterns and facial (00:48:31) expression and tone of voice and so (00:48:33) forth and then if it's in the wrong (00:48:35) hands it could be used to manipulate us (00:48:39) like like never before sure it's our (00:48:41) ultimate vulnerability this beautiful (00:48:43) thing that makes us human becomes this (00:48:47) uh great weakness that we have because (00:48:49) as these AIS continue to self iterate (00:48:54) their capacity to mimic conscious (00:48:57) and human intimacy uh will reach such a (00:49:01) degree of fidelity that it will be (00:49:03) indistinguishable to the human brain and (00:49:04) then humans become like these (00:49:07) unbelievably easy to hack machines who (00:49:10) can be directed wherever the AI you know (00:49:13) chooses to direct them yeah it's not a a (00:49:16) prophecy we we can take actions today to (00:49:19) prevent this uh we can have regulations (00:49:22) about it we can for instance have a (00:49:24) regulation that AIS are welcome to (00:49:26) interact with you humans but on (00:49:28) condition that they disclose that they (00:49:30) are AIS if you talk with an AI doctor (00:49:34) that's good but the AI should not (00:49:36) pretend to be a human being you know I'm (00:49:39) talking with an AI I mean it's not that (00:49:42) there is no possibility that AI will (00:49:44) develop (00:49:45) Consciousness we don't know I mean there (00:49:47) could be that AI will really develop (00:49:50) conscious to such a degree of fidelity (00:49:52) does it even in terms of like how human (00:49:55) beings interact with it does it matter (00:49:57) for the human beings no I mean again (00:49:59) this is the problem I mean because we (00:50:01) don't know if they really have (00:50:02) Consciousness or they're only very very (00:50:04) good at mimicking Consciousness so the (00:50:07) key question is ultimately political and (00:50:09) ethical if they have Consciousness if (00:50:12) they can feel pain and pleasure and love (00:50:15) and hate this means that they are (00:50:17) ethical and political subjects they have (00:50:21) rights that uh you should not inflict (00:50:24) pain on an AI the same way you should (00:50:26) not inflict pain on a human being that (00:50:28) what they like what they love might be (00:50:31) as important as what human beings desire (00:50:34) so they should also vote in elections (00:50:37) and they could be the majority because (00:50:39) you know you can have a country 100 (00:50:41) million humans and 500 million AIS so do (00:50:45) they choose the government in this (00:50:47) situation now you know in the United (00:50:49) States interestingly enough there is (00:50:51) actually an open legal path for AIS to (00:50:54) gain rights it's one of the only (00:50:56) countries in the world where would this (00:50:57) is the (00:50:58) case because in the United States (00:51:00) corporations are recognized as legal (00:51:03) persons with rights until today this was (00:51:07) a kind of legal fiction like according (00:51:09) to US law Google is a person it's not (00:51:12) just a it's a person and as a person it (00:51:14) also have freedom of speech this is the (00:51:17) Supreme Court ruling for 2010 of Citizen (00:51:19) United now until today this was just (00:51:22) legal fiction because every decision (00:51:24) made by Google was actually made by some (00:51:26) human being an executive a lawyer an (00:51:29) accountant Google could not make a (00:51:32) decision independent of the humans but (00:51:34) now you have AIS so imagine the (00:51:37) situation when you incorporate an AI now (00:51:41) this AI is a (00:51:42) corporation and as a corporation US law (00:51:45) recognizes it at a as a person with (00:51:48) certain rights like freedom of speech (00:51:51) now it can earn money it can go online (00:51:54) for instance and offer its services to (00:51:55) people and earn money then it can open a (00:51:58) bank account and invest its money in the (00:52:00) stock exchange and if it's very smart (00:52:02) and very intelligent it could become the (00:52:04) more the richest person in the US now (00:52:07) imagine the richest person in the US is (00:52:09) not a human it's an AI and according to (00:52:12) us slw one of the rights of this person (00:52:16) is to make political contributions (00:52:18) donations this was the main reason (00:52:19) behind citizen United in in (00:52:22) 2010 so this AI now makes billions of (00:52:25) dollars of contributions (00:52:27) to politicians in exchange for expanding (00:52:31) AI (00:52:33) rights so and the legal path is in the (00:52:36) US is completely open you don't need any (00:52:38) new law to make this happen uhhuh that's (00:52:41) like a that's a plot of a (00:52:43) movie yeah when you know we in La yeah I (00:52:46) mean wow that's so wild to contemplate (00:52:50) what are the differences in the ways in (00:52:52) which the Advent of this powerful (00:52:55) technology is impact ing Democratic (00:52:58) systems and authoritarian (00:53:01) systems so both systems have a lot to (00:53:05) gain and have a lot to lose again the AI (00:53:08) it's it's the most powerful technology (00:53:10) ever created it's not a tool it's an (00:53:12) agent so you have millions and billions (00:53:14) of new agents are very intelligent very (00:53:18) capable that can be used to create the (00:53:20) best healthcare system in the world but (00:53:23) also the most lethal army in the world (00:53:26) or the worst secret police in the world (00:53:28) if you think about authoritarian regimes (00:53:31) so throughout history they always wanted (00:53:33) to monitor their citizens around the (00:53:35) clock but this was technically (00:53:37) impossible even in the Soviet Union you (00:53:39) know you have 200 million Soviet (00:53:41) citizens you can't follow them uh all (00:53:45) the time because the the KGB didn't have (00:53:48) 200 million agents and even if the KGB (00:53:51) somehow got 200 million agents that's (00:53:53) not enough because you know in in the (00:53:55) Soviet Union it's still basically paper (00:53:59) bureaucracy the secret police if a (00:54:02) secret agent followed you around 24 (00:54:04) hours a day at the end of the day they (00:54:06) write a paper report about you and send (00:54:09) it to KGB headquarters in Moscow so (00:54:12) imagine every day KGB headquarters is (00:54:14) flooded with 200 million paper reports (00:54:18) now to be useful for anything somebody (00:54:21) needs to read and analyze them they (00:54:23) can't do it they don't have the analysts (00:54:25) therefore even in the Soviet Union some (00:54:28) level of privacy was still the default (00:54:31) for most people uh for technical reasons (00:54:35) now for the first time in history it is (00:54:38) technically possible to annihilate (00:54:40) privacy a totalitarian regime today (00:54:43) doesn't need millions of human agents if (00:54:46) he wants to follow everybody around you (00:54:48) have the smartphones and cameras and (00:54:50) drones and microphones everywhere and (00:54:53) you don't need millions of human (00:54:54) analysts to analyze this o of (00:54:57) information you have ai and this is (00:54:59) already beginning to happen this is not (00:55:01) a future prediction in many places (00:55:04) around the world you begin to see the (00:55:06) formation of this totalitarian (00:55:08) surveillance regime it's happening in my (00:55:10) country in Israel Israel is building (00:55:13) this kind of surveillance regime in the (00:55:14) occupied Palestinian territories to (00:55:17) follow everybody around all the time and (00:55:20) also in our region in Iran since the (00:55:23) Islamic revolution in 1979 they had the (00:55:27) hijab laws which says that every woman (00:55:30) when she goes out walking or even (00:55:33) driving in her private car she must wear (00:55:36) the hijab the head scarve and until (00:55:39) today the regime had difficulty (00:55:43) enforcing the hijab laws because they (00:55:45) didn't have you know millions of police (00:55:48) officers that you can place on every (00:55:50) street a police officer if a woman (00:55:52) drives without a headscarf immediately (00:55:54) she's arrested and fine or whatever in (00:55:57) the last few years they switched to (00:56:00) relying on an AI system Iran Is Now (00:56:03) crisscrossed by uh surveillance cameras (00:56:06) with facial recognition software which (00:56:09) recognizes (00:56:11) automatically if in the car that just (00:56:13) passed by the camera the facial (00:56:16) recognition software can identify that (00:56:19) this is a woman not a man and she's not (00:56:22) wearing the hijab and identify her (00:56:25) identity find her phone number and (00:56:27) within half a second they send her an (00:56:30) SMS message saying you broke the hijab (00:56:33) LW your car is impounded your car is (00:56:35) confiscated stop the car and by the side (00:56:38) of the world this is daily occurrence (00:56:41) today in Teran and isan and other parts (00:56:43) of Iran and uh this is based on AI and (00:56:47) it's not like the there is a report that (00:56:49) go to the court and some human judge (00:56:51) goes over the data and decides what to (00:56:54) do the AI like immediately decides okay (00:56:57) the car is (00:56:59) confiscated and this can happen in more (00:57:01) and more places around around the world (00:57:03) like even in the US you know for for if (00:57:05) you think about all the debate about (00:57:08) abortion without going into the debate (00:57:11) itself the people who think rightly or (00:57:14) wrongly but they think that abortion is (00:57:17) murder they have a very strong incentive (00:57:20) to build a similar surveillance system (00:57:23) for American women you know to stop (00:57:26) murder mhm like you can build this (00:57:28) surveillance system that can identify (00:57:31) yesterday you were pregnant today you (00:57:33) are not what happened in (00:57:35) between so it's not just a problem you (00:57:37) know for Iran or for the Palestinians or (00:57:40) the Chinese this this can come to the US (00:57:43) as (00:57:43) well and to prevent them from crossing (00:57:46) state lines things like that yeah yeah (00:57:49) like okay you went from I don't know (00:57:50) Texas to California you you were (00:57:53) pregnant you came back you're not (00:57:54) pregnant what happened in California so (00:57:57) it feels like AI is this incredible tool (00:58:00) to consolidate power uh around (00:58:02) authoritarian regimes but it also has (00:58:05) its its pitfalls too like it's not the (00:58:07) perfect tool it also frightens the (00:58:10) autocrats uh because the one thing that (00:58:13) human dictators always feared most was (00:58:16) not a democratic Revolution the one (00:58:18) thing they feared most is a powerful (00:58:21) subordinate that they can't control and (00:58:24) that might manipulate them or take power (00:58:26) from them if you can look at the Roman (00:58:28) Empire not a single Roman Emperor was (00:58:31) ever toppled by a democratic Revolution (00:58:34) never happened but many of them uh lost (00:58:38) their life or their power to a (00:58:40) subordinate you know a general that (00:58:43) rebelled against them a provisional (00:58:44) Governor their brother their wife that (00:58:48) took power from them this is the (00:58:50) greatest fear of every dictator also (00:58:53) today and so if you think about AI so if (00:58:56) you're a human dictator and you now give (00:58:59) this immense power to an AI system where (00:59:02) is the guarantee that this system will (00:59:05) not turn against you and either (00:59:08) eliminate you or just turn you into a (00:59:10) puppet I mean what we also know about (00:59:13) dictators it's relatively easy to (00:59:15) manipulate these people if you can (00:59:18) whisper in their ear because they are (00:59:20) very paranoid and the easiest people to (00:59:22) manipulate are the paranoid people and (00:59:25) we have our AI Corporation in the United (00:59:28) States that can deploy billions of (00:59:29) dollars towards Bots and whatever else (00:59:32) to you know create that paranoia or you (00:59:35) really just need to hack one person you (00:59:37) know to to for an AI to take power in (00:59:40) the US very complicated it's such a (00:59:42) distributed system like okay the AI can (00:59:45) learn to manipulate the president but it (00:59:47) also needs to manipulate the Senators (00:59:50) and the Congress members and the state (00:59:52) Governors and the Supreme Court like (00:59:54) what would the AI do with the Senate (00:59:56) phili Buster it's difficult but if you (00:59:58) want to take power in a dictatorship you (01:00:01) just need to learn to manipulate a (01:00:03) single person so uh the dictators are (01:00:06) not all happy about the AIS and we (01:00:10) already beginning to see it for instance (01:00:11) with (01:00:13) chatbots that they are very concerned (01:00:15) because you know you can design a (01:00:18) chatbot which will be completely loyal (01:00:21) to the regime but once you release it to (01:00:25) the internet to start interacting with (01:00:27) people in real life it changes I mean (01:00:31) remember what we talked earlier that AI (01:00:33) is defined by the ability to learn and (01:00:35) change by itself so even if you if Putin (01:00:38) creates like the the Putin's chatbot (01:00:41) that always says that Putin is great and (01:00:43) Putin is right and Russia is great and (01:00:45) so forth but then you release it to the (01:00:47) real world it starts observing things in (01:00:50) the real world for instance it notices (01:00:53) that you know in Russia the invasion of (01:00:56) Ukraine is officially not a war it's (01:00:58) called a special military operation and (01:01:02) if you say that it's a war you go to (01:01:04) prison for up to I think 3 years or (01:01:06) something like that because it's not a (01:01:07) war it's a special military operation (01:01:10) now what do you do if a very intelligent (01:01:12) chatbot That You released you know (01:01:14) connects the dot and says no it's not a (01:01:17) special military operation it's a war (01:01:20) would you send a chat Bo to prison what (01:01:22) what can you do and you know democracies (01:01:25) of course also have a problem with (01:01:27) chatbot saying things we don't like they (01:01:29) can be racist they can be homophobic (01:01:31) whatever but the thing about democracy (01:01:34) it has a relatively wide margin of (01:01:38) Tolerance even for anti-democratic (01:01:41) speech dictatorships have zero margin (01:01:44) for dissenting views so they have a much (01:01:47) bigger problem with how to control these (01:01:50) unpredictable chant (01:01:53) points over the last decade of Hosting (01:01:55) this podcast (01:01:56) my mission has been to engage in what I (01:01:59) consider to be critically important (01:02:01) conversations about the things that (01:02:03) matter most in life while I'm immensely (01:02:05) grateful for the growth of this show (01:02:08) I've also come to realize that my voice (01:02:10) alone is not enough this Mission cannot (01:02:13) be a solitary Endeavor so I wanted to (01:02:16) find a way to help amplify other (01:02:18) meaningful voices and the result is (01:02:21) voicing change media this beautiful (01:02:24) Consortium of thinkers of storytellers (01:02:26) artists and Visionaries all committed to (01:02:29) fostering meaningful exchanges (01:02:31) intentionally curated for those (01:02:33) committed to the path of (01:02:35) self-discovery together we're creating a (01:02:37) space of growth a space of understanding (01:02:40) where every exchange has the potential (01:02:42) to enrich our lives and catalyze (01:02:45) profound personal and planetary change (01:02:49) visit voicing change. media to learn (01:02:51) more And subscribe (01:02:57) how are you interpreting uh the current (01:02:59) moment given that we're on the cusp of (01:03:01) an election here in the United States (01:03:04) and you know there's a lot of discourse (01:03:07) around the existential threat to (01:03:10) democracy that we may be facing uh what (01:03:13) role is AI playing in this what should (01:03:16) we understand about the impact of this (01:03:19) technology on us as Citizens and (01:03:23) voters at present I don't think that AI (01:03:26) has again social media has of of course (01:03:28) a huge impact on the political discourse (01:03:31) and thereby on the results of the (01:03:33) elections but I don't see AI really kind (01:03:36) of changing or manipulating the (01:03:38) elections in November it's too close the (01:03:41) big question is whoever wins the (01:03:44) elections maybe the most important (01:03:47) decisions that person has to make will (01:03:50) be about AI because of the extremely (01:03:53) rapid Pace that this technology is is (01:03:55) developing you know you look at what CH (01:03:58) GPT was a year ago you look at what (01:04:00) things are now in in in 2024 what will (01:04:03) be the state of AI in 2027 (01:04:07) 2028 so you know I watched the (01:04:09) presidential debate most people their (01:04:11) main takeaway was about the cats and the (01:04:13) dogs it's the most memorable thing for (01:04:16) the debate I mean you know whoever wins (01:04:20) maybe we'll have to make some of the (01:04:21) most important decisions in history (01:04:24) about the relations uh I if if you're (01:04:27) worried about immigration it's not the (01:04:29) immigrants that will you know replace (01:04:31) the taxi drivers it's the immigrants (01:04:33) that will replace the bankers that you (01:04:35) should be worried about and it's the AIS (01:04:38) not somebody coming from south of of the (01:04:40) border and who do you trust to make (01:04:44) these momentous decisions now and if you (01:04:47) see think specifically about the threats (01:04:48) to democracy so one thing we learned (01:04:51) from history is that democracies always (01:04:54) since again ancient Athens (01:04:56) they always had this one single big (01:05:01) problem or (01:05:02) weakness that democracy is basically a (01:05:05) kind of a deal that you give power to (01:05:08) somebody for a limited time time period (01:05:11) for four years on condition they give it (01:05:14) back and then you can uh make an a (01:05:17) different Choice like we tried this it (01:05:19) didn't work let's try something else (01:05:21) this ability to say let's try something (01:05:23) else this is democracy and it's B on (01:05:26) that you give power and you expect to (01:05:29) get it back after years transfer at the (01:05:31) end of that term if you give power to (01:05:34) somebody who then doesn't give it back (01:05:37) they now have the power they have the (01:05:40) power to also stay in power that was (01:05:44) always the biggest danger in democracy (01:05:46) so for me the in the issue in the US (01:05:48) elections it's you can discuss the (01:05:50) economic policies the foreign policies (01:05:52) you like this you like that there is (01:05:54) discussion to be had but you have your (01:05:56) one person Donald Trump and that has you (01:05:59) know you have a record from the previous (01:06:02) time that this person doesn't want to (01:06:04) give power back and he is willing to go (01:06:07) a long way including potentially (01:06:09) inciting (01:06:11) violence to uh avoid giving power back (01:06:14) and you want to give him so much power (01:06:17) that doesn't sound like a very a very (01:06:20) good idea so for me this is the kind of (01:06:22) the number one issue in the elections (01:06:24) everything else is is (01:06:26) of marginal importance in comparison (01:06:29) yeah I mean I think it challenges our (01:06:31) our our predels around the stability of (01:06:35) democracy and is forcing us to really (01:06:37) embrace the fact that it is a delicate (01:06:39) Dynamic that is you know informed by (01:06:43) Collective action by the people and in (01:06:47) reflecting upon you know this technology (01:06:50) also uh you know the story of technology (01:06:53) is one in which our ability to legislate (01:06:57) around it and regulate it always falls (01:07:00) you know way behind the pace of (01:07:02) advancement and now we're in a situation (01:07:04) where the pace of advancement is like (01:07:06) nothing we've ever seen before which (01:07:08) calls into question our ability to not (01:07:10) only you know kind of put guardrails (01:07:13) around it but to even understand what is (01:07:15) actually happening the history of (01:07:17) Information Systems is one of collective (01:07:20) human cooperation and yet we're in a a (01:07:23) situation right now where it feels like (01:07:27) cooperation is being challenged not only (01:07:31) nationally here in the United States but (01:07:33) internationally and so as we kind of (01:07:35) begin to talk about how we're going to (01:07:37) triage this or or find Solutions like (01:07:41) where do you land in terms of our (01:07:43) capacity to collectively come together (01:07:47) as a global Community to figure out (01:07:50) Solutions and then put them into motion (01:07:53) so that we don't tiptoe into some kind (01:07:56) of (01:07:56) dystopia so there is a lot to unpack (01:07:59) here so first of all when we think about (01:08:02) cooporation as we said earlier this was (01:08:04) always our biggest Advantage as a (01:08:06) species that we cooperate better than (01:08:08) anybody else we can construct these even (01:08:11) Global networks of trade that no other (01:08:13) animal even understands like if you (01:08:16) think about I don't know (01:08:17) horses so horses never figured out money (01:08:21) they were bought and sold but they never (01:08:23) understood what are these things that (01:08:26) the humans are exchanging and this is (01:08:28) why horses could never unite against us (01:08:32) or could never manipulate us because (01:08:34) they never figured out how the system (01:08:35) works that one person is giving me to (01:08:38) another person in exchange for a few (01:08:40) shiny metal things or some pieces of (01:08:43) paper AI is is different it understands (01:08:47) money better than most people like most (01:08:50) people don't understand how the (01:08:51) financial system really works and (01:08:54) financial AIS inin in Tech they already (01:08:57) surpass most human beings not all human (01:08:59) beings but most human beings in their (01:09:02) understanding of money so we are now (01:09:04) confronting again millions of and (01:09:06) billions of new agents that potentially (01:09:09) can use our own systems against us that (01:09:13) they computers can now collaborate using (01:09:16) for instance the financial system more (01:09:19) efficiently than humans (01:09:21) can so the whole issue of cooporation is (01:09:25) is is changing (01:09:26) and computers also learn how to use the (01:09:29) communication systems to manipulate us (01:09:31) like like in social media so they (01:09:33) cooperating where we are losing the (01:09:35) ability to cooperate and that should (01:09:38) raise the alarm now and the thing that (01:09:41) it's very difficult to understand what (01:09:43) is happening if we want humans around (01:09:46) the world to cooperate on this to build (01:09:49) guard rails to regulate the development (01:09:51) of AI first of all you need humans to (01:09:55) understand what is happening secondly (01:09:57) you need the humans to trust each (01:09:59) other and most people around the world (01:10:02) are still not aware of what is happening (01:10:05) on the AI front you have a very small (01:10:08) number of people in just a few countries (01:10:11) mostly the US and China and a few others (01:10:14) who understand most people in Brazil in (01:10:17) Nigeria in India they don't understand (01:10:21) and this is very dangerous because it (01:10:23) means that a few people many of them are (01:10:25) not even elected by the US ciitizen they (01:10:28) are just you know private companies they (01:10:30) will make the most important (01:10:32) decisions and the even bigger problem is (01:10:35) that even if people start to understand (01:10:37) they don't trust each other like I had (01:10:40) the opportunity to talk to some of the (01:10:43) people who are leading the AI Revolution (01:10:46) which is still led by humans it is still (01:10:48) humans in charge I don't know for how (01:10:49) many more years but as of 2024 it's (01:10:52) still humans in charge and you meet with (01:10:56) these you know entrepreneurs and (01:10:58) business tycoons and politicians also in (01:11:02) the US in China in Europe and they all (01:11:05) tell you the same thing basically they (01:11:08) all say we know that this thing is very (01:11:10) very (01:11:12) dangerous but we can't trust the other (01:11:15) humans if we slow down how do we know (01:11:19) that our competitors will also slow down (01:11:22) whether our business competitors let's (01:11:24) say in here in the US or our Chinese (01:11:27) competitors across the ocean and you go (01:11:29) and talk with the competitors they s the (01:11:31) same thing we know it's dangerous we (01:11:33) would like to slow down to give us more (01:11:35) time to understand to assess the dangers (01:11:38) to debate regulations but we can't we (01:11:41) have to rush even faster because we (01:11:43) can't trust the other Corporation the (01:11:46) other country and if they get it before (01:11:49) we get it it will be a disaster and so (01:11:52) you have this kind of paradoxical (01:11:54) situation (01:11:56) where the humans can't trust each other (01:11:59) but they think they can trust the AIS (01:12:02) because when you talk with the same (01:12:03) people and you tell them okay I (01:12:06) understand you can't trust the Chinese (01:12:07) or you can't trust open AI so you need (01:12:10) to move faster developing the super AI (01:12:13) how do you know you could trust the AI (01:12:16) and then they tell you oh I think that (01:12:18) will be okay I think we've figured out (01:12:20) how to make sure that the AI will be (01:12:23) trustworthy and under our control so you (01:12:26) have this very paradoxical situation (01:12:28) when we can't trust our fellow humans (01:12:31) but we think we can trust and layer on (01:12:33) top of that is an incentive structure of (01:12:35) course that further engenders distrust (01:12:38) in this arms race right like the prize (01:12:40) goes to the Breakthrough developers and (01:12:44) those will be rewarded and remunerated (01:12:47) in ways that are you know perhaps (01:12:49) unprecedented right so absolutely so the (01:12:51) breakthroughs and what's on the other (01:12:53) side of that is is so enticing that any (01:12:57) discourse around regulation or anything (01:12:59) else that might slow it down becomes not (01:13:02) only a national security threat but also (01:13:06) an entrepreneurial threat right so (01:13:07) everything is motivating rapid (01:13:11) acceleration uh at the cost of (01:13:13) transparency and Regulation and all (01:13:15) these other things all these checks and (01:13:16) balances that that we really need right (01:13:18) now and I don't know like you know how (01:13:22) you're feeling about this but it it (01:13:24) leaves me a little cold and and (01:13:26) pessimistic like you're a historian like (01:13:28) the the story of humankind is is all gas (01:13:32) no breaks you know like let's just we're (01:13:35) plowing forward and we'll deal with the (01:13:37) consequences when they come like we're (01:13:39) not wired adequately to really (01:13:42) appreciate the long-term consequences of (01:13:44) our Behavior we're we're kind of you (01:13:46) know looking right in front of us and (01:13:48) making decisions based on how it's going (01:13:50) to impact Us in the immediate future and (01:13:53) and very little else yeah I mean (01:13:55) throughout history the problem is people (01:13:57) are very good at solving problems but (01:13:59) they tend to solve the wrong problems (01:14:02) like they spend very little time (01:14:04) deciding what problem we need to solve (01:14:07) like 5% of the effort goes on choosing (01:14:10) the problem then 95% of the effort goes (01:14:13) in solving the problem we we we focus on (01:14:16) and then we realize oh we actually (01:14:18) solved the wrong problem and it just (01:14:19) creates new problems down the road that (01:14:22) we now need to and then we do it the (01:14:23) same again and you know wisdom often (01:14:27) comes from Silence from taking time from (01:14:31) slowing down let's really understand the (01:14:35) situation before we rush to make a (01:14:38) decision and you know it starts on the (01:14:41) individual level that so many people for (01:14:43) instance think oh my main problem is in (01:14:45) life that is that I don't have enough (01:14:46) money and then they spend the next 50 (01:14:49) years making lots of money and even if (01:14:51) they succeed they wake up at a certain (01:14:53) point and said oops I think I it shows (01:14:55) the wrong problem I think it wasn't yeah (01:14:58) I need some money but it wasn't the my (01:14:59) main problem in life and we are perhaps (01:15:02) doing it collectively as a species the (01:15:04) same thing you know you go back to (01:15:06) something like the Agricultural (01:15:08) Revolution so people thought okay we (01:15:10) don't have enough food let's produce (01:15:13) more food with agriculture we'll (01:15:15) domesticate wheat and rice and potatoes (01:15:17) we'll have lots more food life will be (01:15:19) great and then they domesticate these (01:15:21) plants and also some animals cows (01:15:23) chickens pigs whatever (01:15:26) and they have lots of food and they (01:15:28) start building these huge agricultural (01:15:31) societies with towns and cities and then (01:15:34) they discover a lot of new new problems (01:15:37) they did not anticipate for instance (01:15:39) epidemics hunter gatherers did not (01:15:42) suffer almost any infectious diseases (01:15:45) because most infectious diseases came to (01:15:47) humans from domesticated animals and (01:15:50) they spread in the dense towns and (01:15:52) cities now if you live in a hunter (01:15:54) gatherer band you don't hold any (01:15:56) chickens or pigs so it's very unlikely (01:15:59) some virus will jump from a wild chicken (01:16:02) to you and even if you got some new (01:16:05) virus you have just like 20 other people (01:16:07) in your band and you move around all the (01:16:10) time maybe you infect five others and (01:16:12) like three die and that's the end of it (01:16:15) but once you have these big agricultural (01:16:17) cities then you get the epidemics people (01:16:20) thought they were building Paradise for (01:16:22) humans turned out they were building (01:16:24) Paradise for (01:16:26) germs and human life expectancy and (01:16:28) human living conditions for most humans (01:16:31) actually goes down if you're a king or a (01:16:33) high priest it's okay but for the (01:16:36) average person it was actually a bad (01:16:38) move and the same thing happens again (01:16:41) and again throughout history and it can (01:16:43) happen now on a very very big scale uh (01:16:46) with AI in a way it goes back to this (01:16:49) issue of organic and (01:16:51) inorganic that organic systems are slow (01:16:54) they need time and this AI is an (01:16:57) inorganic system which accelerates (01:16:59) beyond anything we can we can deal with (01:17:02) and the big question is whether we will (01:17:04) force it to slow down or it will force (01:17:07) us to speed up until the the moment we (01:17:10) collapse and die I mean if you force an (01:17:13) organic entity to be on all the time and (01:17:16) to move faster and faster and faster (01:17:19) eventually it collapses and (01:17:20) dies one of the things I heard you say (01:17:23) that that really struck me was (01:17:26) this uh it's a quote if something (01:17:28) ultimately destroys us it will be our (01:17:30) own delusions H so can you elaborate on (01:17:34) that a little bit and how that applies (01:17:36) to what we've been talking (01:17:38) about yeah I mean the AI at least of the (01:17:41) present day they cannot Escape our (01:17:44) control and they cannot destroy us (01:17:46) unless we allow them or unless we kind (01:17:49) of order them to do that we are still in (01:17:53) control but because of our you know (01:17:56) political and mythological delusions we (01:18:00) cannot trust the other humans and we (01:18:03) think we need to develop these AIS and (01:18:06) uh faster and faster and give them more (01:18:09) and more power because we have to (01:18:10) compete with the other humans and this (01:18:12) is the thing that could really destroy (01:18:14) us and you know it's very unfortunate (01:18:17) because we do have a track record of (01:18:20) actually being quite successful of of (01:18:22) building trust between humans it just (01:18:24) takes time (01:18:26) I mean if you think about again the long (01:18:27) Arc of human history so these hunter (01:18:30) gatherer bands tens of thousands of (01:18:33) years ago they were tiny couple of dozen (01:18:36) individuals and even though the next (01:18:39) steps like agriculture they had their (01:18:41) downside again like (01:18:43) epidemics people did learn over time how (01:18:47) to build much larger societies which are (01:18:51) based on trust if you now live in United (01:18:55) States or in some other country you're (01:18:58) are part of a system of hundreds of (01:19:01) millions of people who trust each other (01:19:05) in many ways which were really (01:19:06) unimaginable in the Stone Age like you (01:19:09) don't know (01:19:11) 99.99% of the other people in the (01:19:14) country and still you trust them with so (01:19:18) much I mean the food you eat mostly you (01:19:21) did not go to the forest to hunt and (01:19:23) gather it by yourself you you rely on (01:19:25) Strangers to provide the food for you (01:19:28) most of the tool you use are coming from (01:19:30) strangers your security you rely on (01:19:34) police officers on soldiers that you (01:19:36) never met in your life they are not your (01:19:38) cousins they are not your next door (01:19:40) neighbors and still they protect your (01:19:42) life so yes if you now go to the global (01:19:45) level okay we still don't know how to (01:19:47) trust the Chinese and the Israelis still (01:19:50) don't know how to trust the Iranians and (01:19:52) vice versa but it's not like we are (01:19:54) stuck while we were in the Stone Age (01:19:56) we've made immense progress in building (01:19:59) human trust and we are rushing to throw (01:20:01) it all (01:20:03) away because uh it just again it takes (01:20:06) time it will not happen tomorrow yeah I (01:20:09) mean I think it's urgent that we find a (01:20:10) way back to repairing some institutional (01:20:14) trust right like that has been degraded (01:20:16) in recent times and I think without that (01:20:20) uh we stand very little chance as a (01:20:24) democratic Republic of surviving and (01:20:27) solving these kinds of problems (01:20:29) absolutely if if you ask in brief what (01:20:33) is the key to building trust between (01:20:35) millions of strangers the key is (01:20:37) institutions because you can't build a (01:20:39) personal intimate relationship with (01:20:42) millions of people so it's only (01:20:44) institutions whether it's courts or uh (01:20:47) police forces or newspapers or (01:20:50) universities or healthc Care Systems (01:20:52) that build trust between people (01:20:56) and unfortunately we now see this uh (01:20:59) again another epidemic of distrust in (01:21:02) institutions on both the right and the (01:21:04) left it is fueled by a very cynical (01:21:08) worldview which basically says that the (01:21:10) only reality is power and humans only (01:21:14) want power and all human interactions (01:21:17) are power (01:21:18) struggles so whenever somebody tells you (01:21:20) something you need to ask whose (01:21:23) privileges are being served (01:21:25) whose interests are being Advanced and (01:21:28) any institution is just a elite (01:21:30) conspiracy to take power from us so (01:21:33) journalists are not really interested in (01:21:35) knowing the truth about anything they (01:21:37) just want power and the same for the (01:21:39) scientists and the same for the judges (01:21:42) and if this goes on then all trust in (01:21:45) institutions collapses and then Society (01:21:47) collapses and the only thing that can (01:21:49) still function in that situation is a (01:21:51) dictatorship because dictatorships don't (01:21:53) need trust they are based on terror so (01:21:56) people who attack institutions they (01:21:58) often think oh we are liberating the (01:22:01) people from these authoritarian (01:22:03) institutions they are actually Paving (01:22:06) the way for a (01:22:07) dictatorship and the thing is that this (01:22:10) view is not just very cynical it's also (01:22:13) wrong humans are not these power crazy (01:22:17) demons all of us want power to some (01:22:19) extent that's true but that's not the (01:22:21) all truth about us humans are really (01:22:23) interested in knowing the the truth (01:22:25) about ourselves about our lives about (01:22:28) the world on a very deep level because (01:22:30) you can never be happy if you don't know (01:22:33) the truth about your life are because (01:22:36) you will not know what are the sources (01:22:38) of misery again you will focus on your (01:22:40) life if you don't know the truth you (01:22:42) waste all your life trying to solve the (01:22:45) wrong problems and this is true of also (01:22:48) of journalists and judges and scientists (01:22:52) yes there there is corruption in every (01:22:54) Institution this is why we need a lot of (01:22:56) Institutions to keep each one another in (01:22:59) check but if you destroy all trust in (01:23:02) institutions what you get is either (01:23:05) Anarchy or a (01:23:07) dictatorship and again it's a good (01:23:09) exercise every now and then to stop and (01:23:11) think about how every day we are (01:23:14) protected by all kinds of Institutions (01:23:17) like when people talk with me about the (01:23:19) Deep State you know this conspiracy (01:23:21) about the Deep State I immediately think (01:23:23) about the sewage system (01:23:26) the sewage system is the Deep State it's (01:23:28) a deep H system of tunnels and pipes and (01:23:33) pumps which is the state built under our (01:23:37) houses and streets and neighborhoods and (01:23:40) saves our life every day because it (01:23:42) keeps our sewage separate from our (01:23:45) drinking water you know you go to the (01:23:47) toilet you do your thing it goes down (01:23:49) into the deep state which keeps it (01:23:52) separate from the drinking water (01:23:55) uh if I can tell one historical anecdote (01:23:57) where did it come from so you know after (01:24:00) Agricultural Revolution you have big (01:24:02) cities they are Paradise for germs hot (01:24:05) beds for epidemics this continues really (01:24:08) until the 19th century London in the (01:24:11) 19th century was the biggest city in the (01:24:12) world and one of the most dirty and (01:24:15) polluted and a hot bed for epidemics and (01:24:17) in the middle of the 19 century there is (01:24:19) a cholera epidemic and people in London (01:24:22) are dying from cholera and then you have (01:24:24) this bureaucrat medical bureaucrat Jon (01:24:26) Snow not the guy from Game of Thrones a (01:24:29) real Jon Snow who did not fight dragons (01:24:32) and zombies but actually did save (01:24:35) millions of lives cuz he went around (01:24:38) London with lists and he interviewed all (01:24:41) the people who got sick or who died if (01:24:44) somebody died from Colorado he would (01:24:45) interview their family tell me where did (01:24:48) this person get their drinking water (01:24:50) from and he made these long lists of (01:24:53) hundreds and thousands of people and by (01:24:56) analyzing these lists he pinpointed a (01:24:59) certain well on Broad Street in SoHo in (01:25:02) London where everybody almost everybody (01:25:05) who got sick on colera they had a zip of (01:25:08) water from that well at a certain stage (01:25:11) and he convinces the municipality to (01:25:13) disable the pump of the of the well and (01:25:16) the epidemic stops and then they (01:25:18) investigate they discover that the well (01:25:20) was dug about a meter away from a (01:25:23) cesspit and one water sewage water from (01:25:26) the cesspit got into the drinking water (01:25:29) and today if you want to dig a well or a (01:25:31) cesspit in London or in Los Angeles you (01:25:34) have to fill so many forms and to get (01:25:36) all these bureaucratic permits and it (01:25:38) saves our lives and how does that relate (01:25:41) to this idea of the deep state I'm (01:25:43) trying to tether those two Notions (01:25:45) together again the people who believe (01:25:47) the conspiracy theories about the Deep (01:25:48) State they say that all all these State (01:25:51) bureaucracies they are Elite conspiracy (01:25:54) is against the common people trying to (01:25:57) take over power trying to destroy us and (01:26:00) in most cases no the people in this you (01:26:03) know to manage a seage system you need (01:26:06) plumbers you also need bureaucrats again (01:26:09) you need to apply for a license to dig a (01:26:11) well and it is managed by all these kind (01:26:14) of state bureaucrats and it's a very (01:26:16) good thing because again there is (01:26:18) corruption in these places sometimes (01:26:21) this is why we keep also courts you can (01:26:23) go to court this this is why we keep (01:26:25) newspapers so they can expose corruption (01:26:28) in the cities in the municipalities (01:26:30) sewage department but most of the time (01:26:34) most of these people are honest people (01:26:37) who are working very hard every day to (01:26:40) keep our sewage separate from our (01:26:42) drinking water and to Keep Us Alive and (01:26:44) by extrapolation there are all of these (01:26:46) bureaucracies that are working in our (01:26:48) interest in invisible ways that we take (01:26:50) for granted exactly basically right (01:26:52) you've often said Clarity is power power (01:26:55) and I think your superpower is your (01:26:57) ability to kind of stand at 10,000 ft (01:26:59) and look down on Humanity in the planet (01:27:02) and (01:27:04) identify what's most important in these (01:27:07) macro trends that help us make sense of (01:27:11) what's Happening Now and I'd like to (01:27:13) kind of end this with some thoughts on (01:27:16) how you cultivate that clarity through (01:27:19) meditation and your you know very kind (01:27:22) of like profound uh practice of (01:27:24) mindfulness and information deprivation (01:27:27) I should say right yeah information (01:27:30) fasts yeah starting maybe is with the (01:27:33) idea of an information fast so I think (01:27:36) this is important today for every person (01:27:39) to go in an information diet that this (01:27:42) idea that more information is always (01:27:44) good for us it's like thinking that more (01:27:46) food is always good for us it's it's not (01:27:47) true and the same way that the world is (01:27:50) full of junk food that we better avoid (01:27:53) the world is also full of junk (01:27:55) information that we have better avoid (01:27:58) information which is (01:27:59) artificially filled with greed and hate (01:28:03) and fear information is the food of the (01:28:05) mind and we should be as mindful as what (01:28:09) we put into our minds as of what we put (01:28:11) into our mouths but it's not just about (01:28:14) limiting (01:28:15) consumption it's also about digesting (01:28:18) it's also about (01:28:20) detoxifying like we go throughout our (01:28:23) life and we take in a lot of junk (01:28:26) whether we like it or not that fills our (01:28:28) mind and I I meditate two hours every (01:28:31) day so I can tell you there is a lot of (01:28:33) junk in there a lot of hate and fear and (01:28:38) greed that I picked up over the years (01:28:41) and it's important to take time to (01:28:44) Simply digest the information and to (01:28:47) also detoxify to kind of let go of all (01:28:51) this hatred and and anger and fear and (01:28:54) and uh and greed which is in our (01:28:56) minds so I began when I was doing my PhD (01:28:59) in Oxford a friend recommended that I go (01:29:02) on a Meditation Retreat or vasana a (01:29:05) meditation and for a year he kind of (01:29:07) nagged me to go on and I said no this is (01:29:09) kind of mystical mambo jumbo I don't (01:29:11) want to to to and eventually I went and (01:29:14) it was amazing because it was the most (01:29:16) remote thing for mysticism that I could (01:29:19) imagine uh because I it was a 10 days (01:29:23) Retreat and on the very first evening of (01:29:25) the retreat the teacher Essen goenka the (01:29:28) only instruction he gave he didn't tell (01:29:30) me to kind of visualize some godess so (01:29:33) do this man nothing he just said what is (01:29:36) really happening right now bring your (01:29:39) attention to your nostrils to your nose (01:29:42) and just feel whether the breath is (01:29:44) going in or whether the breath is going (01:29:47) out that's the only exercise like a pure (01:29:52) observation of reality what amazed me (01:29:55) was my inability to do it like I would (01:29:58) bring my attention to the nose and try (01:30:00) to feel is it going in is it going out (01:30:02) and after about 5 Seconds some thought (01:30:05) some memory some fantasy would arise in (01:30:07) the mind and would just hijack my (01:30:09) attention and for the next two or three (01:30:11) minutes I would be rolling in this (01:30:14) fantasy or memory until I realize hey I (01:30:17) actually need to observe my breath and I (01:30:19) would come back to the Breath Again 5 (01:30:21) seconds maybe 10 seconds I will be able (01:30:23) oh now it's coming in it's coming in oh (01:30:25) now it's going out it's going out and (01:30:27) again some memory would come and hijack (01:30:29) me and I realized first that I've I know (01:30:32) almost nothing about my mind I have no (01:30:35) control of my mind and my mind is just (01:30:38) like this Factory that constantly (01:30:41) produces fantasies and Illusions and (01:30:44) delusions that come between me and (01:30:47) reality like if I can't observe the (01:30:50) breath going in and out of my nostrils (01:30:53) because some fantasy comes up what hope (01:30:56) do I have of understanding AI or (01:31:00) understanding the conflict in the Middle (01:31:01) East without some mindmade illusion or (01:31:05) fantasy coming between me and (01:31:08) reality and for the last 24 years I have (01:31:11) this daily exercise of I devote two (01:31:13) hours every day to just what is really (01:31:16) happening right now I sit with closed (01:31:19) eyes and just try and focus let go of (01:31:22) all all the mindmade stories and feel (01:31:26) what is happening to the breath what is (01:31:28) happening to my body the reality of the (01:31:31) present moment I also go for a long (01:31:34) Meditation Retreat usually every year of (01:31:36) between 30 days and 60 days of (01:31:38) meditation uh because again one of the (01:31:41) things you realize there is so much (01:31:42) noise in the mind that just to calm it (01:31:46) down to the level that you can really (01:31:48) start meditating seriously it takes (01:31:51) three or four days of continuous (01:31:53) meditation (01:31:55) just so much noise so long Retreats they (01:31:59) enable to have this really deep (01:32:01) observation of reality which is (01:32:04) impossible most of life we spend like (01:32:07) detached from reality two hours a day (01:32:11) that's a commitment even in the midst of (01:32:13) all the book promotion craziness you're (01:32:17) able to find came here I I usually do (01:32:19) one in the morning one in the afternoon (01:32:21) or evening what a beautiful thing and (01:32:23) obviously your ability to think clearly (01:32:26) and write so articulately about these (01:32:29) ideas is very much a product of this (01:32:32) practice absolutely I mean without the (01:32:35) practice I would not be able to write (01:32:36) such books and I would not be able to (01:32:38) deal with the kind of all the publicity (01:32:41) and all the interviews and you know this (01:32:43) roller coaster of positive and negative (01:32:46) feedback from the world all the time I (01:32:49) would say one one important thing this (01:32:51) is not necessarily for everybody (01:32:54) because I meditate and I have meditator (01:32:56) friends and so forth I mean different (01:32:58) things work for different people there (01:33:00) are many people that I wouldn't (01:33:02) recommend to meditate two hours a day or (01:33:05) to go for a 10 days Meditation Retreat (01:33:08) because they are different their body (01:33:09) their minds are different for them (01:33:11) perhaps going on a 10 days hike in the (01:33:14) mountains would be better for them (01:33:17) perhaps devoting two hours a day to (01:33:19) music to to say playing or to creating (01:33:22) or going to to psychotherapy y would (01:33:25) have better results humans are really (01:33:27) different in many ways from one another (01:33:29) there is no one size fits all so if you (01:33:32) never try meditation absolutely try it (01:33:35) out and and and give it a real chance (01:33:37) it's not like you go for like a few (01:33:39) hours and it doesn't work okay give it (01:33:41) up like give it a real chance but keep (01:33:43) in mind that again different minds are (01:33:46) different um so find out what really (01:33:48) works for you and whatever it is that's (01:33:51) the important part whatever it is invest (01:33:53) in it (01:33:55) I have to release you back to your life (01:33:58) uh but maybe we can end this with just a (01:34:01) a concise thought about what it is that (01:34:02) you want people to take away from from (01:34:04) this book like what is most vital and (01:34:07) crucial for people to understand about (01:34:09) what you're trying to (01:34:11) communicate but information isn't truth (01:34:14) truth is a it's it's a costly a rare and (01:34:18) precious thing it is the foundation of (01:34:21) of knowledge and wisdom and of nine (01:34:24) beneficial societies you can build (01:34:27) terrible societies without the truth but (01:34:29) if you want to build a good society and (01:34:31) you want to build a good personal life (01:34:33) you must have a a strong basis in the (01:34:37) truth and it's difficult again because (01:34:39) most information is is not the truth and (01:34:43) invest in it it's worthwhile uh to have (01:34:46) a practice whatever it is that gets you (01:34:49) connected with reality that gets you (01:34:51) connected with the truth thank you for (01:34:54) for coming here today uh I really (01:34:56) appreciate you taking the time to share (01:34:58) your wisdom and experience I think uh (01:35:00) Nexus your latest book is as I said at (01:35:03) the outset a crucial vital book that (01:35:06) everybody should read uh we're entering (01:35:08) into a very interesting time and we are (01:35:12) well advised to be as best prepared as (01:35:14) we possibly can and uh I appreciate the (01:35:17) work that you do um and thank you again (01:35:20) you've all thank you I only graced the (01:35:22) surface of the outline that I cre so (01:35:24) hopefully you can come back CU I got a (01:35:25) million more questions I could have (01:35:27) talked to you for hours next time I'm in (01:35:29) La I'll be happy to thanks man (01:35:31) appreciate it cheers (01:35:37) peace that's it for today thank you for (01:35:40) listening I truly hope you enjoyed the (01:35:42) conversation to learn more about today's (01:35:44) guest including links and resources (01:35:47) related to everything discussed today (01:35:49) visit the episode page at Rich roll.com (01:35:52) where you can find the entire podcast (01:35:55) archive my books Finding Ultra voicing (01:35:58) change in the plant power way as well as (01:36:00) the plant power meal planner at meals. (01:36:03) roll.com if you'd like to support the (01:36:05) podcast the easiest and most impactful (01:36:08) thing you can do is to subscribe to the (01:36:10) show on Apple podcast on Spotify and on (01:36:14) YouTube and leave a review and or (01:36:16) comment this show just wouldn't be (01:36:18) possible without the help of our amazing (01:36:20) sponsors who keep this podcast running (01:36:23) wild and free to check out all their (01:36:25) amazing offers head to Rich roll.com (01:36:28) slss sponsors and sharing the show or (01:36:31) your favorite episode with friends or on (01:36:33) social media is of course awesome and (01:36:36) very helpful and finally for podcast (01:36:38) updates special offers on books the meal (01:36:41) planner and other subjects please (01:36:43) subscribe to our newsletter which you (01:36:44) can find on the footer of any page at (01:36:47) Rich roll.com Today's show was produced (01:36:49) and engineered by Jason Cameo the video (01:36:52) edition of the podcast was created by (01:36:54) Blake Curtis with assistance by our (01:36:56) creative director Dan Drake portraits by (01:36:59) Davey Greenberg graphic and social media (01:37:01) assets courtesy of Daniel siss and thank (01:37:04) you Georgia Wy for copywriting and (01:37:06) website management and of course our (01:37:08) theme music was created by Tyler Patt (01:37:10) Trapper Patt and Harry Mattis appreciate (01:37:13) the love love the support see you back (01:37:16) here soon peace plance namaste (01:37:21) [Music]

Leave a Reply

Your email address will not be published. Required fields are marked *