AI and the future of humanity | Yuval Noah Harari at the Frontiers Forum

foreigner, thank you for this wonderful introduction and yes, what I want to talk to you about is AI in the

future

of

humanity

. Now I know that this conference focuses on the ecological crisis facing

humanity

, but for better, for worse, ai2 is part of this crisis. AI can help us in many ways to overcome the ecological crisis or it can make it much worse, in fact, AI will probably change the very meaning of the ecological system because for 4 billion years the ecological system of planet Earth contained only organic life forms and We could now or soon see the emergence of the first inorganic life forms within 4 billion years or at least the emergence of inorganic agents.

Now people have feared AI since the beginning of the computer age in the mid-20th century and this fear has inspired many science fiction classics such as Terminator from The Matrix, although these science fiction scenarios have become cultural landmarks, have not generally been taken seriously in academic, scientific and political debates and perhaps for good reason because science fiction scenarios generally assume that Before AI can pose a significant threat to humanity, it will have to reach or overcome important milestones. First, AI will have to become sentient and develop Consciousness, feelings, emotions, otherwise why would it want to take over the world?

Secondly, the AI ​​will have to become Adept. To navigate the physical world, robots will have to be able to move and operate in houses, cities, mountains and forests at least as dexterously and efficiently as humans, if they can't move through the physical world, how can they take over it? April 2023 AI still seems to be far from reaching any of these milestones despite all the hype around GPT chat and the other new AI tools, there is no evidence that these tools have even a shred of consciousness, feelings of emotions when it comes to navigating the physical world despite the commotion.

Regarding autonomous vehicles, the date when these vehicles will dominate our roads continues to be postponed; However, the bad news is that, to threaten the survival of human civilization, AI doesn't really need Consciousness and it doesn't need the ability to move around the world. In the physical world, new AI tools have been released in the public sphere in recent years that may threaten the survival of human civilization from a very unexpected direction and it is difficult for us to even understand the capabilities of these new AI tools and In fact, they continue to develop because AI is capable of learning on its own to improve itself, even the developers of these tools do not know all the capabilities of what they have created and are often surprised by the skills and qualities emerging from these tools.

I assume everyone here already knows some of the most fundamental abilities of the new AI tools, such as writing text, drawing images, composing music, and writing code, but there are many additional capabilities that are emerging, such as deepfaking the voices and images of people, like drafting bills, finding weaknesses. both in computer code and in legal contracts and agreements, but perhaps most importantly, new artificial intelligence tools are gaining the ability to develop deep and intimate relationships with human beings. Each of these skills deserves a full discussion and is difficult for us to understand. All its implications, so let's keep it simple when we take all these skills together as a package, they boil down to one very big thing: the ability to manipulate and generate language, whether with words, images or sounds, the most important aspect of today .

The phase of the AI ​​revolution underway is that AI is gaining mastery of language at a level that surpasses the average human ability and by gaining mastery of language, AI is taking the master key that opens the doors of all our institutions, from banks to temples, because language is the tool we use to give instructions to our bank and also to inspire heavenly visions in our minds. Another way to look at it is that AI has just hacked the operating system of human civilization. The operating system of all human cultures in history has always been language.

In the beginning was the word, we use language to create mythology and laws to create gods and money to create Art and Science to create friendships and Nations, for example, human rights are not a biological reality, they are not inscribed in our DNA, human rights are something We create with language by telling stories and writing laws. The gods are also not a biological or physical reality. God is also something that we humans have created with language by telling legends and writing scriptures. Money is not a biological or physical reality. The bills are simply worthless pieces. paper and today more than 90 percent of the money in the world is not even bills, it is just electronic information on computers that is passed around, what gives value to money of any kind is only the stories of people like bankers, finance ministers and cryptocurrency gurus. tell us about the money Sam Bankman freed Elizabeth Holmes and Bernie Madoff didn't create much real value, but unfortunately they were all extremely capable storytellers.

What would it mean for human beings to live in a world where perhaps most stories are melodies, images, laws? The policies and tools are shaped by a non-human alien intelligence that knows how to exploit with superhuman efficiency the weaknesses, prejudices and addictions of the human mind and also knows how to establish deep and even intimate relationships with human beings; That is the big question today in In games like chess, no human being can hope to beat a computer, what would happen if the same thing happened in art, politics, economics and even religion? When people think about Chan GPT and other new AI tools, they are often drawn to examples like children using Chan GPT to write their school essays.

What will happen to the school system when children write essays with horrible shajipiti but these types of questions miss the big picture forget about school essays instead think about, for example, the next US presidential race in 2024 and try to imagine the impact of new artificial intelligence tools that can be mass produced political manifestos, fake news and even sacred writings for new sects. In recent years, the politically influential Q Anon sect has formed around anonymous online texts known as Q, there are now millions of followers of this sect. in the US, in the rest of the world, they revised and interpreted these Q drops as a kind of new writing, so the sacred text now, as far as we know, all the previous Q drops were composed by human beings and the Bots only helped spread these texts. online, but in the

future

we could see the first cults and religions in history whose revised texts were written by a non-human intelligence and of course religions throughout history claimed that their sacred books were written by a non-human intelligence.

This was never true before this. could come true very, very quickly, with far-reaching consequences. Now, on a more prosaic level, we might soon find ourselves carrying on long online discussions about abortion or about climate change or about the Russian invasion of Ukraine with entities we think are human beings, but who are in fact AI Bots. Now the problem is that it is completely useless, there is no point in wasting our time trying to convince an AI bot to change its political opinions, but the more time we spend talking to the bot, the better it knows us and understands how to do it. hone your messages to change our political opinions or our economic opinions or anything else through your command of language.

AI, as I said, could also form intimate relationships with people and use the power of intimacy to influence our opinions and well See now There is no indication that AI has, as I said, any consciousness, feelings of its own, but to create false intimacy with human beings, AI doesn't need feelings of its own, it just needs to be able to inspire feelings in the US to catch us. which will now be attached in June 2022, was a famous incident when Google engineer Blake Lemon publicly claimed that the AI ​​chat over Lambda he was working on had become sensitive.

This highly controversial statement cost him his job. What was really interesting about this episode wasn't Lemwan's claim, which was probably false, what was really interesting was his willingness to take a chance and ultimately lose his very lucrative job to the AI ​​chatbot he thought he was protecting if the AI can influence people to take risks. and losing his jobs, what else can he induce us to do in every political battle for hearts and minds? Intimacy is the most effective weapon of all and AI has just acquired the ability to produce intimacy with millions, hundreds of millions of people now, as we probably all know that over the last decade social media has become a battlefield. , a battleground for controlling human attention.

Now, with the new generation of AI, the battlefront is shifting from attention to intimacy and this is very bad news about what will happen to human society and human psychology. AI fights AI in a battle to create intimate relationships with us, relationships that can then be used to convince us to buy certain products or vote for certain politicians. Even without creating false intimacy, new AI tools would have an immense influence on human opinions and in our vision of the world that people, for example, can use, are already beginning to use a single AI advisor as a comprehensive oracle and as a source of all the information they need, it is not surprising that Google Be terrified if you've been watching Google's newsletter. you are terrified and at a good pace, why bother searching for yourself when you can just ask the Oracle to tell you whatever you want, you don't need to search the news industry and the advertising industry should be terrified too, right?

Why read a newspaper when I can? just ask the Oracle to tell me what's new and what's the point, what's the purpose of the ads, when I can ask the Oracle to tell me what to buy, so there's a chance that in a very short time the entire industry advertising collapse. While I or the people and companies that control the new AI oracles will become extremely powerful, what we are potentially talking about is nothing less than the end of human history, not the end of History, just the end of the human-dominated part of what we call history is the interaction between biology and culture, it is the interaction between our biological needs and desires for things like food and sex, and our cultural creations, such as religions and laws, history is the process through which religions and laws interact with food and sex.

What will happen to the cause of this story interaction when AI takes over the culture in a few years. AI could eat up all human culture. Everything was produced for thousands and thousands of years. To eat it all, digest it and begin to sprout an avalanche of new things. Cultural creations new cultural artifacts and remember that we humans never have direct access to reality, we are always enveloped by culture and we always experience reality through a cultural prison, our political opinions are shaped by the stories of journalists and by the anecdotes of friends. of sexual preferences are modified by movies and fairy tales, even the way we walk and breathe is something driven by cultural traditions now, before, this cultural cocoon was always woven by other human beings, tools that helped spread , previous tools such as printing presses, radios or televisions. cultural ideas and creations of humans, but they could never create something new by themselves.

A printing press cannot create a new book, a human AI always does. It is fundamentally different from the radio printing presses of all previous inventions in history because it can completely create. new ideas can create a new culture and the big question is what it will be like to experience reality through a prism produced by a non-human intelligence by an alien intelligence now, at first, in the early years, AI will probably largely imitate the prototypes. the human prototypes that nurtured it in its infancy, but with each passing year, AI culture will boldly go where no human has gone before, so for thousands of years we humans basically live within dreams and fantasies. of other humans,We have worshiped gods and pursued ideals. beauty we dedicate our lives to causes that originated in the imagination of some poet, prophet or human politician, we could soon find ourselves living within the dreams and fantasies of an alien in intelligence and danger that has potential danger, it also has positive potential. but the dangers he presents are fundamental, very different from all or most of the things imagined in science fiction movies and books.

Previously, people primarily feared the physical threat posed by intelligent machines, so Terminator depicted robots running through the streets and shooting people. The Matrix It was assumed that to gain full control of human society, AI would first need to gain physical control of our brains and connect them directly to the computer network, but this is incorrect by simply gaining mastery of human language. The AI ​​has everything you need to make coconuts. a matrix as a world of illusions, contrary to what some conspiracy theories assume, it is not really necessary to implant chips in people's brains to control or manipulate them for thousands of years.

Poets and politicians have used language and storytelling to achieve profit. manipulating and controlling people and reshaping society is now likely to be done by AI and once it can doubt that it doesn't need to send killer robots to shoot us, it can make humans pull the trigger if it really needs to. So far, the fear of AI has haunted humanity only for the last few generations, let's say since the middle of the 20th century, if we go back to Frankenstein, maybe it's 200 years, but for thousands of years humans have been haunted by a much deeper fear.

I always appreciated the power of stories and images and language to manipulate our minds and create illusions consequently since ancient times humans feared being trapped in a world of illusions in the 17th century when decout feared that perhaps a malicious demon was trapping him inside of this type of world of illusions creating everything that the cult saw and heard in ancient Greece Plato told the famous allegory of the cave in which a group of people are chained inside a cave all their lives in front of a blank wall a screen on that screen they see projected several Shadows and the error of the prisoner these Illusions these Shadows For the reality in ancient India, the Buddhist and Hindu wise men pointed out that all humans lived trapped within what they called Maya.

Maya is the world of illusions. Buddha said that what we normally consider reality are often just fictions in our own minds, people can take entire actions. Wars kill others and they are willing to die themselves because of their belief in these fictions, so the AI ​​revolution is bringing us face to face with the card demon, with Plato's cave and with the Mayans If we are not careful, a curtain of illusions could descend over the world. all of humanity and we will never be able to tear that curtain off or even realize that it is there because we will think that this is reality and social media if this sounds crazy, so just look at social media in recent years.

The media has given us a small sample of what is to come in social networks primitive artificial intelligence tools Artificial intelligence tools, but very primitive, have been used not to create content but to select content produced by human beings, humans produce stories, videos and whatever, and artificial intelligence. choose which stories, which videos would reach our ears and eyes, select those that will receive the most attention, that will be the most viral and, although they are very primitive, these artificial intelligence tools have been enough to create this type of curtain of illusions that increased social polarization everywhere.

The world undermined our mental health and destabilized democratic societies. Millions of people have confused these illusions with the reality that the United States has the most powerful information technology in all of history, and yet American citizens can no longer agree on who won the last election or whether the change climate. is real or whether vaccines prevent diseases or not, the new AI tools are much more powerful than these social media algorithms and could cause much more damage now, of course, AI also has huge positive potential. I didn't talk about it because we people who Develop AI naturally talk about it enough, there's no need for me to join that chorus.

The job of historians and philosophers like me is often to point out the dangers, but certainly AI can help us in countless ways, from finding new cures for cancer to discovering solutions to the ecological crisis we face to ensuring that new AI tools are used for good and not evil, we must first appreciate their true capabilities and must regulate them very carefully since 1945. We knew that nuclear technology could physically destroy human civilization, in addition to benefiting us by producing cheap and abundant energy. Therefore, we reform the entire international order to protect ourselves and ensure that nuclear technology is used primarily for good.

Now we have to deal with a new weapon of mass destruction that can annihilate our mental and social world and a big difference between nuclear weapons and AI. Nuclear weapons cannot produce more powerful nuclear weapons. AI can produce more powerful AI, so we must act quickly before AI gets out of our control. Pharmaceutical companies can't. sell people new medicines without first subjecting these products to rigorous safety checks. Biotech labs can't simply release a new virus into the public sphere to impress their shareholders with technological wizardry; Similarly, governments must immediately prohibit the release into the public domain of any more revolutionary viruses.

AI Tools Before They Are Safe Again I'm not talking about stopping all research in AI, the first step is stopping its release in public school. Is there somewhere you can research viruses without releasing them to the public. You can research AI, but. let's not release them too quickly into the public domain. If we don't stop the AI ​​arms race, we won't have time to even understand what's happening, let alone effectively regulate this incredibly powerful technology. You may be wondering or wondering. We want to slow down the public deployment of AI because democracies are lagging behind the most ruthless retirement regimes and the answer is absolutely no, exactly the opposite, unregulated deployment of AI is what will cause democracies to lose to dictatorships, because if we unleash chaos, authoritarian regimes could more easily contain this chaos than open societies could democracy in essence is a conversation democracy is an open conversation you know that dictatorship is a dictation there is a person who dictates everything no conversation democracy is a conversation between many people about what to do and conversations are based on language when AI hacking language means it could destroy our ability to carry out meaningful public conversations, thus destroying democracy if we wait for chaos It will be too late to regulate it democratically, perhaps in an authoritarian regime a totalitarian rate will still be possible to regulate, but how?

Can you regulate something democratically if you can't have the conversation about it and if you didn't regulate AI in time you won't be able to? We will no longer be able to have a meaningful public conversation, so in conclusion, we have basically met. an extraterrestrial intelligence not in outer space, but here on Earth we don't know much about this extraterrestrial intelligence except that it could destroy our civilization, so we should stop the irresponsible deployment of this extraterrestrial intelligence in our societies and regulate AI before it happen. regulates us and the first regulation that we could suggest in many regulations, but the first regulation that I would suggest is that it is mandatory for the AI ​​to disclose that it is an AI if I am having a conversation with someone and I cannot tell if it is an AI. a human being or an AI.

That is the end of democracy because it is the end of meaningful public conversations. What do you think about what you just heard in the last 20 or 25 minutes? I guess some of you might be alarmed, some of you might be angry. with the corporations that develop these Technologies or the governments that do not regulate them, some of you may be angry with me thinking that I am exaggerating the threat or the time to mislead the public, but whatever you think, I bet my words have had something of Emotional impact on you, not just intellectual impact, also emotional impact.

I just told you a story and this story is likely to change your mind about certain things and may even cause you to take certain actions in the world. Who created this story? I just listened and that just changed your mind and your brain. Now I promised you that I wrote the text of this presentation myself with the help of some other humans, although the images were created with the help of AI. I promise. Do you know that at least the words you heard about the cultural product of one human mind or several human minds, but can you be absolutely sure that this is the case?

A year ago, a year ago there was nothing on Earth, at least not in the public domain other than a human mind that could produce such a sophisticated and powerful text, but now it is different, in theory, the text you just heard it could have been generated by a non-human alien intelligence, so take a moment or more to think about it. Thank you, that was an extraordinary presentation and I'm actually going to find out how many of you found that terrifying. There are a lot of very smart people here who found that terrifying. There are many questions to ask.

I'm going to take some from the audience and some online, so a gentleman here and the editor of the Borders and Sustainability tour was a wonderful presentation. I love his book. I follow it with all my heart, so one question among many is that Regarding the regulation of AI that regulates AI, I very much agree with the principle, but now the question is how correct it is, so I think It is very difficult to build the nuclear reactor in your basement, but you can definitely train your AI in your basement quite easily. So how can we regulate that?

One kind of question related to that is that this whole Frontiers Forum and the

forum

is really about open science, open information, open data and most of the AI ​​that exists is trained using publicly available information, including patterns, books and scriptures , so regulating AI doesn't mean that we should regulate and put that information into a confined space, which goes against the open science and open data initiatives that we're also really thinking are really important to us. The black box is an algorithm, right? That's the algorithm, no, I know there will always be trade-offs and the question is simply to understand what kind of regulations we need, first we need time now, nowadays, these artificial intelligence tools are very Powerful technologies are not yet produced by individual hackers in their basements, they need a huge amount of computing power, they need a huge amount of money, so it's being led by just a few major corporations and governments and, again, it's going to be very, very difficult to regulate something on a global level because it is an arms race, but there are things that countries have the benefit of regulating even just themselves, like again this example of an AI, when it interacts with a human being, it must reveal that it is an AI, even if some authoritarian regime does not.

If you want to do it, the US EU or other democratic countries can have this and this is essential to protect the open society. Now there are a lot of questions around online censorship, so you have this controversy about Twitter or Facebook, who authorized them? For example, preventing the former president of the United States from making public statements and this is a very complicated issue but there is a very simple problem with the boss, you know that human beings have freedom of expression. Bots don't have enough expression, it's a human right. I have it, but I don't, so if you deny freedom of speech to robots, I think that should be fine with everyone, uh, let's take another question: if you could pass the microphone here.

I am Prince, the most beloved and I am a philosopher. I just have an interesting question, oh, I think that's an interesting question, there you go. I have a question for you regarding your choice of language to go from artificial to extraterrestrial because artificial suggests that there is still some kind of human control, whereas I believe it is extraterrestrial. suggests something strange, but it also suggests atless in the imagination of a way of life, so I'm curious to know what work you're trying to make those words do for you, yes, it's definitely still artificial in the sense that we produce it, but it's becoming more so. by producing itself, it learns and adapts more and more on its own, so artificially that it is a kind of illusion that is still under our control and is slipping out of our control, so in this sense it is becoming in an alien force, not necessarily evil again, can also do a lot. of good things, but the first thing we have to realize is that it's extraterrestrial, we don't understand how it works or in one of the most shocking things about all this technology, you talk to the people who run it and you ask them questions about how it works.

Can I do it and it says they said we don't know? I mean, we know how we built it initially, but then it really learns on its own. Now there is a whole discussion about whether it is a way of life or not. I think he doesn't have Consciousness yet and uh, I don't think it's impossible for him to develop Consciousness, but I also don't think it's necessary for him to develop Consciousness, that's a problem, it's an open question, but not life. It does not necessarily mean consciousness, we have many forms of life, microorganisms, plants, any fungi that we believe they do not have.

Consciousness, we still consider them a way of life and I think AI is getting very close to that position, ultimately, of course, what is it? life is a philosophical question, I mean, we define the boundaries and you know, whether it's a virus, life or not, we think that an amoeba is life, but a virus is somewhere right on the boundary between life and non-life. life, then it's, you know, it's Language is our choice of words, so I think it's less important, of course, what we call AI, but the most important thing is to really understand what we're dealing with and not console ourselves with this kind of illusions.

Oh, it's something we created. It is under our control if it does something wrong we will just disconnect it, no one knows how to disconnect it anymore. I'm going to answer a question from our online audience, um, this is from Michael Brown in the US. What do you think about the possibility that artificial general intelligence already exists and that she or those who have access to artificial intelligence generally already influencing social systems? I think it's very, very unlikely that we wouldn't be sitting here if artificial general intelligence really existed. When I look at the world and the chaotic scenario, I mean that artificial general intelligence is really the end of human history and it is something so powerful that it is not something that anyone can contain and so when I look at the chaotic state of the world, I'm pretty sure again, from a historical perspective, that no one has it anywhere.

How long will it take to develop artificial general intelligence? I don't know, but to threaten the foundations of civilization we don't need general artificial intelligence and then move on. Going back to social media, a very, very primitive AI was still enough to create huge social networks and political chaos if I think about it in evolutionary terms, so AI is now just coming out of the organic soup like the first organisms that came out of the organic soup four billion years ago, how long will it take for Tyrannosaurus Rex to reach it? it will take to reach Homo sapiens not until a billion years it could be only 40 years that is the thing about digital evolution it is moving on a completely different time scale to organic evolution I can thank you it has been absolutely wonderful it has been a great pleasure have you here And I have no doubt that you will stay with us for a while afterward, but everyone in the audience, please join me in saying thank you.