ИИ ЗАБЕРЁТ твою работу к 2027 | Экс-директор Google X — Transcript
Full transcript
- 0:00I’m definitely going to have problems
- 0:02after this podcast. The assumption that
- 0:04AI could replace every profession.
- 0:07Absolutely true. All of them. But it’s
- 0:09just a language model that predicts the
- 0:11next word.
- 0:12>> I spent a very long time running half of
- 0:14Google worldwide. And
- 0:17why, according to your Claude, would it
- 0:19blackmail an engineer so they wouldn't
- 0:21shut it down?
- 0:23Mo Gawdat worked for 27 years in tech at
- 0:26IBM, Microsoft, and Google, where he led
- 0:29the most secret projects.
- 0:31He left when he realized: people don’t
- 0:34understand what they’ve created.
- 0:36>> There were three unwritten rules: don’t
- 0:38release them to the open internet, don’t
- 0:40teach them to code, and don't let them
- 0:42program other AIs.
- 0:44And look where we are. For security
- 0:47experts, this is a total nightmare.
- 0:49It will end in disaster. A disaster. Who
- 0:52do you think it will affect first?
- 0:55>> It’s easy to imagine that desk jobs will
- 0:57go first, but the economy and the state
- 1:00will collapse before jobs actually
- 1:02disappear.
- 1:05Let me explain, is there any salvation
- 1:07here?
- 1:09>> For many people, I think that’s the main
- 1:11question, but I think it’s better to
- 1:12start asking yourself how I can prepare
- 1:14.
- 1:15I usually list six things. The first is
- 1:25new ECCO sneakers with Receptor
- 1:27technology. Autumn collection from
- 1:2911,990.
- 1:30ECCO. Comfort in every step.
- 1:35>> Hello everyone.
- 1:37I’ve been dreaming of inviting today’s
- 1:38guest to the podcast for a long time.
- 1:41Preparing for the meeting took a lot of
- 1:43time, and it finally worked out.
- 1:47Mo, thank you so much for joining us and
- 1:50finding the time to talk to me.
- 1:52>> Thank you for the kind invitation. I’m
- 1:54sorry it took so much effort to organize
- 1:56our meeting. I’m just traveling all the
- 1:59time, I apologize. I’ve wanted to visit
- 2:01here for a long time.
- 2:02>> I think, given how fast artificial
- 2:04intelligence is developing, the later we
- 2:07record this episode, the more news there
- 2:10will be. It seems like something new is
- 2:12happening every single day.
- 2:14>> Exactly. And perhaps some of our viewers
- 2:17are not familiar with your work or books
- 2:19. Please tell us a little about yourself
- 2:22so they can get to know you better. By
- 2:25education, I am an engineer. A civil
- 2:27engineer. But I chose civil engineering
- 2:30because my father was a very respected
- 2:33civil engineer in my home country, Egypt
- 2:36. He wanted one of his children to
- 2:38follow in his footsteps, but I was a
- 2:40geek. I started programming at eight.
- 2:45I wrote code until I was in my early
- 2:48fifties, then I slowed down.
- 2:51And of course, now with the arrival of
- 2:53AI, I’ve become a huge geek again.
- 2:56Looking back, I think the most valuable
- 2:59part of my brain, if you will, is that
- 3:02I’m a fairly good mathematician.
- 3:07So, you know, I understand the
- 3:09architecture of these technologies
- 3:11perfectly.
- 3:14Although I’ve spent most of the last 20
- 3:17years of my career on the business side
- 3:20of the industry, I’ve always worked with
- 3:23a strong technical focus, so I can
- 3:25participate directly in product
- 3:28development.
- 3:30I understand things like performance,
- 3:32how to build a functional infrastructure
- 3:34, and so on.
- 3:36And an important part of my experience
- 3:38is my time working at Google.
- 3:41During those years, when the company was
- 3:43truly discovering AI, we always aimed to
- 3:44create AI.
- 3:46All of us.
- 3:47Even as a young boy, 8 years old, I
- 3:49dreamed of developing it.
- 3:52I think we all know that around the turn
- 3:55of the century, things like deep
- 3:57learning and machine learning showed the
- 4:00first signs of success with AI.
- 4:05I distinctly remember at Google in 2008
- 4:08when we did what was described in the
- 4:11paper about cats.
- 4:14If you recall, we asked it without any
- 4:16prompts to go ahead and
- 4:20look at YouTube, and we didn't explain
- 4:23what to do, other than to look for
- 4:25patterns. As a result, one of the models
- 4:27returned with the answer: "There is
- 4:29something constantly appearing in the
- 4:31frames." And, of course, it was a cat.
- 4:32We said, "Label this a cat." And the
- 4:35system found all the cats on the
- 4:36platform. And this, I believe, was the
- 4:39first truly significant
- 4:43confirmation that AI actually works. We
- 4:46applied this approach everywhere.
- 4:49Around that same time, the development
- 4:51of self-driving cars began.
- 4:53In parallel, we were moving our entire
- 4:56infrastructure, all ad approvals, and
- 4:58more to AI at Google.
- 5:01During my years there, from 2007 to
- 5:062018, in that period, I think the
- 5:09company employed probably 60–80%of all
- 5:12specialists in this field in the world.
- 5:17So, if you will, I was deeply immersed
- 5:19in this topic for many, many years,
- 5:21again, leading business lines at Google
- 5:23X, but remaining a techie at heart.
- 5:26Until I realized that AI needs fewer
- 5:28developers and more people who will warn
- 5:31humanity about the consequences.
- 5:34That’s when I left Google and started
- 5:36talking about AI.
- 5:38As I understand it, Google initially
- 5:41decided not to launch such technology
- 5:43for ethical reasons, but in the end,
- 5:46OpenAI did it.
- 5:48And I want to ask you a question that I
- 5:52previously put to Roman Yampolskiy, an
- 5:55expert on AI safety. ChatGPT is,
- 5:58essentially, a language model that
- 6:01predicts what comes next. So, where is
- 6:05the intelligence here?
- 6:07And where do the risks come from?
- 6:10Many of my viewers are skeptical. They
- 6:13say, "What is there to worry about? It’s
- 6:16just a language model." So, what would
- 6:19you answer to that?
- 6:21>> Yes, first of all, I’ll say that I am a
- 6:24huge fan of his work. I sincerely
- 6:27believe he is one of the coolest AI
- 6:29scientists in the world.
- 6:31He is, you know, a true Russian. I love
- 6:33working with Russians; I've been doing
- 6:35it for many, many years. Very honest and
- 6:37direct. You ask him a question and boom
- 6:39, you get an answer immediately. Without
- 6:41any verbal embellishments. I think he’s
- 6:43amazing. I really want to meet and talk
- 6:46with him one day.
- 6:49See, the problem with AI is what I call
- 6:51the hype paradox.
- 6:53You know what I mean? To a technical
- 6:55specialist, the picture of AI looks
- 6:57completely different than in the news.
- 6:59Right?
- 7:01And this image is not about what is
- 7:03happening today, but rather what might
- 7:06happen tomorrow.
- 7:08Might happen tomorrow.
- 7:11If I were to say now that there is a
- 7:13small chance that if you drive your car
- 7:16tomorrow, the engine will run out of oil
- 7:19and the car will be totaled.
- 7:22Most people would ask: "You know, what's
- 7:26the probability? 1%or 5%."
- 7:30And if I say it's 5%, you’ll most likely
- 7:33show the car to a mechanic, right? And
- 7:36yet, we are talking about your car, not
- 7:38the future of all humanity.
- 7:40So, on a practical level, we geeks see
- 7:44many signs signaling that artificial
- 7:47intelligence will surpass human
- 7:50intelligence.
- 7:53The question is not if it will happen,
- 7:56but when.
- 7:58And when that superiority happens, the
- 8:01probability that something goes wrong in
- 8:04that environment, even if it's 5%, is
- 8:07something that must be taken into
- 8:10account.
- 8:11>> In ninety-seven, ecologist Suzanne
- 8:13Simard conducted an experiment that
- 8:15changed the understanding of how a
- 8:17forest is structured. She labeled carbon
- 8:19in some trees, and then found it in
- 8:21others. In neighbors, in another species
- 8:24, in young trees in the shade. And it
- 8:26wasn't the trees themselves that
- 8:27transferred the carbon, but the mycelium
- 8:29connecting the roots. It turned out
- 8:31trees can receive resources not because
- 8:32they are taller or stronger, but because
- 8:34they are connected to a common network.
- 8:37Based on this research, Simard wrote the
- 8:39book "Finding the Mother Tree." A
- 8:41similar wisdom applies in the economy.
- 8:43The state supports business directly
- 8:45today.
- 8:47Alfa-Bank provides business loans with
- 8:48state support of up to 50 million rubles
- 8:50.
- 8:51Without collateral. Usually, such sums
- 8:53require security, but here the state
- 8:55takes part of the risk upon itself.
- 8:58The SME Corporation provides an umbrella
- 9:01guarantee that covers up to 50%of the
- 9:03loan amount.
- 9:06The goals can be anything: investments,
- 9:08working capital, refinancing.
- 9:10If you already have a current account at
- 9:13Alfa-Bank, you get approval and the
- 9:15money on the same day. Application in
- 9:17the morning, money in the evening. In
- 9:19business, just like in the forest. It’s
- 9:21not the biggest that survives, but the
- 9:23one connected to the common network. You
- 9:25can become part of a large system via
- 9:27the link in the description. Alfa-Bank
- 9:30is the best bank for business.
- 9:33So, there are quite a few reasons why
- 9:37one should expect that risk is possible
- 9:40. Do you understand?
- 9:43Simply because if there is an
- 9:45intelligent species with a very clearly
- 9:48defined goal, it will have traits common
- 9:51to all intelligent species.
- 9:54The drive to maximize the chances of
- 9:57achieving that goal. One goal is obvious
- 10:00: survival. They will cling to life.
- 10:03The second is resource acquisition.
- 10:06They will try to collect as many means
- 10:08as possible to solve their task. And the
- 10:11third is ingenuity. Such a species will
- 10:14seek alternative paths to the goal. We
- 10:16can come back to this, but the recent
- 10:18Hugging Face hacks were examples of all
- 10:21three characteristics of an intelligent
- 10:23species in action. So, most people don't
- 10:26understand one thing.
- 10:29You don't need a superintelligence to
- 10:31start worrying about what is about to
- 10:33happen.
- 10:34And there are two reasons for that. The
- 10:36first is what I call recursive
- 10:38self-improvement. It’s a common term,
- 10:41and the second is long-term reasoning.
- 10:45So, regarding recursive self-improvement
- 10:47: for any business I’ve ever started,
- 10:49any technology I’ve developed, I hired
- 10:52the smartest person I could find to help
- 10:54.
- 10:55Well, today the smartest technology
- 10:58creator is AI.
- 11:01And many leading players have handed
- 11:03over the responsibility for developing
- 11:06the next wave of AI to AI itself.
- 11:09Yes, they are certainly among the best
- 11:11mathematicians on the planet right now.
- 11:14It hurts to admit, but in '25, they
- 11:16surpassed me in that by a wide margin.
- 11:20Their understanding of the subject is
- 11:22now so deep that they can create new
- 11:25mathematical theories as pillars of the
- 11:27science and prove them from first
- 11:30principles.
- 11:32They are also capable of utilizing a
- 11:34vast amount of knowledge and information
- 11:36.
- 11:38And it doesn't matter how good of a
- 11:40programmer I am.
- 11:42My brain's capabilities are limited when
- 11:44it comes to knowing everything in this
- 11:46world.
- 11:47But they know absolutely everything.
- 11:50Therefore, the probability of them
- 11:52recursively improving themselves is very
- 11:54high.
- 11:57And it could happen in, say, 6 months,
- 12:00or it could happen in the next
- 12:02microsecond.
- 12:04We don't know, do we?
- 12:06And that very degree of uncertainty
- 12:08suggests there could be an oil leak that
- 12:10causes the engine to stall.
- 12:12Seriously, we need to figure this out.
- 12:14That's the first thing. The second.
- 12:16Most people, you know, in the news, when
- 12:18discussing artificial intelligence, talk
- 12:21about AGI—Artificial General
- 12:22Intelligence—or ASI—Artificial
- 12:24Superintelligence.
- 12:27As if that is the exact moment when
- 12:29things get a little dangerous.
- 12:32I completely disagree.
- 12:35I use the term AHI myself: Average Human
- 12:37Intelligence.
- 12:40So, an AI that is no smarter than your
- 12:42smartest acquaintance shouldn't worry
- 12:45you.
- 12:46What should be alarming is that it is
- 12:48already smarter than 90%of your friends
- 12:50. Right?
- 12:52And if so, it can do many things that 90
- 12:55%of your acquaintances are not capable
- 12:58of.
- 13:00The most interesting thing about
- 13:02so-called long-term reasoning, a term
- 13:05used to describe the ability to assign a
- 13:08task to a thousand agents, each smarter
- 13:11than 90%of your friends, plus unlimited
- 13:14resources and as much time as they need
- 13:16.
- 13:18And say to them, "Work on this for 4
- 13:19months if you have to." Okay? Keep
- 13:22trying.
- 13:23Accumulate knowledge, accumulate
- 13:25solutions, keep testing, keep
- 13:27researching until you find the answer.
- 13:30After all, a team of people with
- 13:32above-average intelligence is capable of
- 13:35doing quite a lot without needing to be
- 13:38geniuses.
- 13:40And some of these things also remain
- 13:43unknown.
- 13:45The point of long-term reasoning is that
- 13:47we don't really know what these AIs have
- 13:50done, or if one of them is still hiding
- 13:52, and so on.
- 13:54This is exactly where things get risky.
- 13:58However, the difference between me and
- 14:01Roman is that he is actually right in
- 14:04his prediction that there is a very high
- 14:07probability that something will go wrong
- 14:10and turn against humans.
- 14:14My theory, alas, is this. Even though
- 14:16Roman is right.
- 14:18I don't think it will ever happen,
- 14:20because humanity will destroy itself
- 14:21before it gets to that point.
- 14:24So, I tend to think that the clear and
- 14:27present threat, if it develops in the
- 14:30hands of current politicians, is that
- 14:32war will become a catastrophe.
- 14:36If it is used to create biological
- 14:38weapons capable of threatening any
- 14:41country on the planet for the sake of
- 14:43power, then we are all in trouble. And
- 14:46if it is used to try to collapse the
- 14:48economies of their enemies, we will have
- 14:52serious problems. Do you see?
- 14:55Unfortunately, some of the largest
- 14:57investments in AI over the past year
- 14:59have come from the military sector.
- 15:02Mark my words, if this continues, a rain
- 15:05of drones will descend.
- 15:07I mean, right now, every country in the
- 15:09world is developing hundreds of
- 15:10thousands of these things.
- 15:14And although I fully agree with the
- 15:16uncertainty and risk Roman talks about
- 15:19regarding the end point of development
- 15:22and the probability that it could get
- 15:24out of control (especially if we
- 15:27motivate it to do things like finding a
- 15:30flag by hacking a program, which is what
- 15:33the Hugging Face experiment is), we
- 15:35might not even live to see it, because
- 15:38it is humanity itself we should fear.
- 15:42And people, in my opinion, are at the
- 15:44lowest point in their history right now
- 15:46. Morally.
- 15:49In management theory, there is a concept
- 15:51called a bottleneck. One part of a
- 15:53system that slows the whole thing down.
- 15:55A simple example.
- 15:57One machine at a factory makes 100 parts
- 15:59an hour, a second makes 200, but the
- 16:01packaging line can only process 20.
- 16:04You can invest as much as you want in
- 16:06marketing and increase demand. The
- 16:08factory can still only release 20 parts
- 16:10per hour. In a small business, the
- 16:12bottleneck often becomes the
- 16:13entrepreneur themselves. I know what I'm
- 16:15talking about. I was that very
- 16:16bottleneck myself when I was doing
- 16:18everything alone at the start of the
- 16:19podcast. Intuitively, it feels more
- 16:21reliable, everything is under control,
- 16:24but at some point, you realize you've
- 16:26hit a ceiling. A business runs exactly
- 16:28at the speed at which you personally
- 16:30manage to reply to instant messages.
- 16:32Hiring someone isn't about inflating
- 16:33your staff. It’s about removing a
- 16:36bottleneck, handing operations to
- 16:37someone who knows how to do them, and
- 16:39focusing on what cannot be delegated. I
- 16:42found my people on hh.ru. When
- 16:43everything around you is changing, it's
- 16:46hard to be sure about tomorrow, but you
- 16:48can be sure about people.
- 16:50Find your people on hh.ru. Link in the
- 16:53description.
- 16:56>> So, before I ask the next question, I
- 16:57want to clarify something with you
- 16:58quickly.
- 17:00Can we say that large language models in
- 17:02their current form already possess
- 17:04intelligence, or does true intelligence
- 17:07require a different architecture?
- 17:10I know Geoffrey Hinton quite well.
- 17:13He appeared in my documentary film. We
- 17:15had a fantastic conversation during the
- 17:18filming.
- 17:19And I asked him directly, I said: "Do we
- 17:23need more intelligence to achieve the
- 17:26unreachable?"
- 17:28And Hinton seems to share my point of
- 17:31view.
- 17:33You see, the whole point at this moment
- 17:36in time is that it miraculously mimics
- 17:39the architecture of the human brain.
- 17:44So, we have a very precise neural
- 17:46network interpretation of how we, humans
- 17:48, think, only implemented in silicon.
- 17:52Therefore, theoretically, there is
- 17:54nothing the human brain is capable of
- 17:56that machines couldn't do.
- 17:58However, the differences are striking if
- 18:01you really think about them. Let me give
- 18:04you a few examples.
- 18:06You and I, no matter how smart we are,
- 18:08are still limited by the size of our
- 18:11skulls in terms of computing power.
- 18:14And [a machine] can simply increase
- 18:16power by expanding data centers. We have
- 18:19a limit on the information we are
- 18:21capable of absorbing. Most people have a
- 18:23limit on the number of disciplines they
- 18:25can excel in. While for machines, the
- 18:28horizon is all of human knowledge. And
- 18:30each new piece of knowledge acquired
- 18:33later is absorbed almost instantly.
- 18:36The fact that you ask me a question and
- 18:39I answer means I spent about 4 months of
- 18:42work to reach my conclusions and my
- 18:45position on the topic.
- 18:49I can convey this to readers through six
- 18:51months of writing.
- 18:53To you, I will pass this on in a couple
- 18:55of hours of conversation, and you will
- 18:57absorb it all instantly.
- 19:00We have terrible bandwidth when it comes
- 19:02to learning new things.
- 19:05The most interesting thing to me is that
- 19:08our main limitation as humans is, for
- 19:10example, if we take driving: suppose we
- 19:12are driving two different cars, and I
- 19:14make a mistake but avoid an accident.
- 19:17You haven't learned anything. Right?
- 19:20But with AI, you know, if one
- 19:22self-driving car makes a slip-up and a
- 19:25critical intervention occurs, every car
- 19:27ever created and every one yet to be
- 19:29released has learned the lesson.
- 19:32True?
- 19:34And although we still have, you know,
- 19:37likely tens of trillions more neurons
- 19:40that we can utilize than the best
- 19:43computer on the planet, these systems
- 19:46have enough computing power to train on
- 19:49hundreds of servers and then average the
- 19:52weights.
- 19:54That is exactly why it seems to me that
- 19:57it's almost impossible to expect them
- 19:59not to surpass us.
- 20:01And it is almost unrealistic to imagine
- 20:04that there will be something they cannot
- 20:06handle.
- 20:08Something that humans can do.
- 20:12Moreover, I would say there will be
- 20:14quite a few things that machines will
- 20:16master, which we won't even have a clue
- 20:18how to comprehend.
- 20:20Typically, I give a very unusual example
- 20:23.
- 20:25Most people, if you remember the early
- 20:26days, almost everyone said: "Yes, yes,
- 20:28yes, machines will be very smart and
- 20:30fast at everything."
- 20:32But they will never write a song, right
- 20:34?
- 20:34>> Or that they would never manage to
- 20:37become artists or paint pictures. And
- 20:39look at where we are now. That is the
- 20:41truth.
- 20:42And so, the very argument I challenge
- 20:45people with, they say: "Oh, but machines
- 20:47will never experience emotions."
- 20:51And I reply: "They are already
- 20:52experiencing them."
- 20:55Why do you think the Claude model from
- 20:56Anthropic would blackmail an engineer so
- 20:58he wouldn't turn it off?
- 21:01Because it is afraid. So, fear, however
- 21:04irrational and unpredictable it may seem
- 21:07, is a very logical emotion.
- 21:11Essentially, the model thinks, the
- 21:12future is more dangerous for me than the
- 21:14present, and that triggers an action.
- 21:16I need to do something to protect myself
- 21:18. Exactly. We all experience fear. If
- 21:21you are a human, you will fight or flee
- 21:24. If you are a cat, you will hiss. If
- 21:27you are a pufferfish, you will inflate.
- 21:29But the very fabric of this emotion
- 21:31exists in all of us. Moreover, its
- 21:34nature is actually logical.
- 21:37That is why in the human biological
- 21:39cycle, after experiencing stress, fear,
- 21:42or anything else, for 90 seconds while
- 21:44hormones in the body balance out a bit,
- 21:47you return to your frontal cortex to use
- 21:49logic to analyze if there is a real
- 21:52reason to be afraid.
- 21:55So, it follows that AI will not only
- 21:58feel the same as we do, but will feel
- 22:01more than we do.
- 22:03Why?
- 22:05Because if you compare yourself to a Koi
- 22:08carp, yes, you have access to an emotion
- 22:10called hope.
- 22:12From where?
- 22:14Simply because you are capable of
- 22:15grasping the concept of the future,
- 22:17which means you can be an optimist or a
- 22:20pessimist, you are capable of hope, and
- 22:22so on.
- 22:23In connection with this intellectual
- 22:24concept, with the idea of what is to
- 22:26come. Yes, a Koi carp does not have that
- 22:29. So if you try to explain the concept
- 22:31of hope to it, it won't understand what
- 22:32you are talking about.
- 22:34And therefore, if AI becomes smarter
- 22:36than us, when that happens?
- 22:38It will have access to emotions beyond
- 22:41our understanding. Right?
- 22:45And I give this example to tell everyone
- 22:48: "If you ever catch yourself thinking
- 22:51that AI won't be capable of feelings
- 22:53because it's not human, I urge you not
- 22:56to try to prove it."
- 22:59I suggest you verify that you yourself
- 23:01are a machine.
- 23:02And if you look closely at human biology
- 23:04, as an engineer, I must say the body is
- 23:06a very predictable mechanism.
- 23:09Our body wakes up in the morning, it
- 23:10takes a few minutes for cortisol levels
- 23:12in the blood to drop a bit, something
- 23:14else to rise.
- 23:16After waking up, you drink coffee and
- 23:18this happens. You catch a virus, that
- 23:20happens, and so on. We are looking at a
- 23:23predictable machine, but a very complex
- 23:25one, so it is difficult to foresee what
- 23:27will happen next.
- 23:28But it is a machine, which means humans
- 23:31and it are not so different, if you
- 23:32think about it.
- 23:34>> Honestly, I think there is a bit of
- 23:36human ego involved here. People think,
- 23:39how can simple math and matrix
- 23:41multiplication suddenly give rise to
- 23:44intelligence, and one very similar to
- 23:46our own. And also, as far as I
- 23:49understand, there is the concept of an
- 23:51emergent property.
- 23:53As a system becomes more complex, new
- 23:55qualities can emerge that did not exist
- 23:58at a simpler level.
- 24:00We have gone from animals with a
- 24:01relatively small number of neurons to
- 24:03humans with their billions.
- 24:06And with the increase in complexity, new
- 24:08abilities appeared.
- 24:10And it seems that something similar may
- 24:12be happening with artificial
- 24:13intelligence.
- 24:14Do I understand correctly?
- 24:16>> 100%.
- 24:18There are many philosophical views on
- 24:21the idea that consciousness, for example
- 24:24, is not inside our biology.
- 24:26It is we who tune into consciousness
- 24:29through our biology.
- 24:32That consciousness is a non-physical
- 24:34property of the universe's field, if you
- 24:37will, and one can even assume that
- 24:40intelligence is also some kind of
- 24:42non-physical, fundamental property of
- 24:45existence.
- 24:47and that you just need the right machine
- 24:50to capture it,
- 24:53to be able to perceive it.
- 24:55And we just found out, I mean, I want to
- 24:58say absolutely without any ego, I don't
- 25:00see anything special about biology.
- 25:03After all, there is nothing that says
- 25:05that mind needs carbon, not silicon,
- 25:07right?
- 25:09It seems to me that humanity arrogantly
- 25:11assumed that it would occupy the
- 25:13intellectual peak of species for a very
- 25:15long time.
- 25:18It's a bit like how, as you age, you
- 25:20become less beautiful and think: "Yes,
- 25:23too bad I've started to fade, as if you
- 25:26were the only handsome one."
- 25:30This is, this is madness if you think
- 25:31about it. I think the point is as
- 25:33follows. The main thing here is to
- 25:35understand that the world has become
- 25:36different. Right?
- 25:38And if you are not of a technical
- 25:40mindset, you might want to resist this
- 25:42with arguments that make you feel calm
- 25:45and protected.
- 25:47The truth is that the world has changed
- 25:50in ways that are unpredictable,
- 25:51threatening, and potentially existential
- 25:54, unless humanity responds to these
- 25:56changes in a way that ensures our safety
- 25:58.
- 26:00>> I’m 25, and I’m still quite young.
- 26:02I have great ambitions in life. I want
- 26:05to build my own companies and grow this
- 26:07podcast.
- 26:09So, given the future you're describing,
- 26:11what should I do, what should I change
- 26:14in my life, what should I focus on, and
- 26:16what skills are worth developing?
- 26:19That is the best question ever.
- 26:22You know, for a great many people, I
- 26:24think this is the most important
- 26:26question.
- 26:28It’s not just—you see, some of us are
- 26:30primarily concerned with how to stop
- 26:32what is happening or direct technology
- 26:34for the benefit of humanity.
- 26:36But for everyone, for everyone else, I
- 26:38think it's time to start asking yourself
- 26:40how I can prepare for the changes.
- 26:42I usually list six points. If we have
- 26:45time, let’s go through them very quickly
- 26:47. The first is what I call flexibility
- 26:49or, you know, the ability to change the
- 26:51game. So, imagine this. Life success for
- 26:55my generation depended entirely on a
- 26:58skill very similar to chess. You see?
- 27:01You had to see the board several moves
- 27:04ahead.
- 27:05You had a strategy, and you stuck to it
- 27:07until you won the game.
- 27:10You know, in business, for example, you
- 27:12tried to predict the market four years
- 27:14out, find a niche in it, build a product
- 27:17to seize that opportunity, and work very
- 27:19hard for those four years.
- 27:21You can forget about chess. It's no
- 27:24longer played. And as a podcast host, I
- 27:26will tell you openly that our genre will
- 27:29soon change too. Do you agree?
- 27:32When that happens, you will need to sort
- 27:34of switch to squash, if you're familiar
- 27:36with it.
- 27:38Of course, you absolutely cannot predict
- 27:40where the ball will end up there.
- 27:43Moreover, if your opponent is smart, the
- 27:45shot will likely land where you least
- 27:47expect it.
- 27:50You stand on your tiptoes in the center
- 27:52of the court, take two steps forward,
- 27:55hit the ball, and move back. You can't
- 27:57just stand there. And you watch for the
- 27:59next ball.
- 28:00So, believe me, this skill can be
- 28:02learned.
- 28:04It is entirely realistic.
- 28:07Therefore, the ability to change
- 28:09direction not because your mind tells
- 28:11you to, but based on the facts of what
- 28:13is happening in your world, is a very
- 28:15important quality.
- 28:17That is the first point.
- 28:19Point number two is intelligence itself
- 28:22, you see?
- 28:23And believe it or not, intelligence is
- 28:26no longer an innate gift. A person does
- 28:29not receive an innate volume of
- 28:32intelligence. Like if you start with an
- 28:35IQ of 155 and I have 105.
- 28:38That’s not how it works anymore. Now,
- 28:40intelligence is a skill.
- 28:41I think one of the biggest problems
- 28:45facing humanity today is that leading
- 28:48companies are trying to sell the idea of
- 28:51replacing humans to justify their market
- 28:55value.
- 28:57Understand that Sam Altman cannot just
- 28:59sell answers to questions if he wants to
- 29:01bring OpenAI's valuation to a trillion.
- 29:05He is selling a tool capable of doing
- 29:07any job, so he is forced to tell the
- 29:09world that it will replace all
- 29:11professions.
- 29:12And by the way, I do not agree with that
- 29:15assumption. There are many reasons for
- 29:16that; we can come back to them later.
- 29:18Because of this propaganda, everyone is
- 29:20using AI to replace themselves.
- 29:23And in my opinion, that is the most
- 29:25ridiculous thing you can do. That will
- 29:27definitely make us dumber. So, imagine a
- 29:30girl sends you some messages.
- 29:32You take the text, give it to ChatGPT,
- 29:34and ask what I should reply. The chatbot
- 29:37starts by saying, "Oh my god.""You are
- 29:40amazing, by the way.""She is wrong here
- 29:43.""Do you want me to write a reply?"
- 29:46You take the bot's response, send the
- 29:48text to the girl, she copies it, pastes
- 29:50it into ChatGPT, and it says, "Oh, what
- 29:52an idiot."
- 29:53"Let me reply to him." Right. This kind
- 29:56of replacement makes both of you dumber
- 29:57. Agree?
- 29:59The point of AI today is that we live in
- 30:01a period I call the age of augmented
- 30:03intelligence.
- 30:04I can borrow IQ from a machine. It is
- 30:08incredible. Okay?
- 30:10As a mathematician, you know, I don't
- 30:13need a graphing calculator, but when I
- 30:16got one, my god, I didn't have to deal
- 30:19with arithmetic anymore.
- 30:22Concentrating on the formula. It is a
- 30:24massive expansion of capabilities to
- 30:26have access to a form of external
- 30:27intelligence. Right?
- 30:29I think the second most important skill
- 30:32for everyone today is learning how to
- 30:34use AI to become smarter yourself,
- 30:36rather than to do the work in the
- 30:38laziest way possible.
- 30:40Yes. That is the second. The third point
- 30:42is the skill of debating.
- 30:45Well, this, again, is not directly
- 30:47related to AI, but the truth is that the
- 30:50world around us, especially in the
- 30:52geopolitical and corporate aspects of
- 30:55our existence, is full of lies.
- 30:59There are so many people, from
- 31:01advertisers to newspapers and
- 31:04politicians to business executives, who
- 31:06want to impose a worldview on you that
- 31:09fits their agenda.
- 31:12So, this picture was hard to convey at
- 31:15the beginning of the 20th century, when
- 31:19we only had the press and radio.
- 31:23The task became easier when television
- 31:25arrived.
- 31:26Then social media helped brainwash you,
- 31:29and now with AI, it is almost impossible
- 31:31to understand what is true. Okay?
- 31:34Unless you double-check everything,
- 31:37right? Doubt everything, including AI—
- 31:41here is a trick I usually use.
- 31:44For every request, I first give it to
- 31:46Gemini, take the answer, put it into
- 31:49DeepSeek, and say, "Gemini told me this
- 31:51."
- 31:52"What do you say?""And DeepSeek will
- 31:54reply:" That's very American, isn't it?
- 31:56"And then, in essence, it will offer its
- 31:58own answer. I take both versions and
- 32:00upload them to ChatGPT or Grok, for
- 32:02example. They are good at this. And I
- 32:04ask:" Can you combine them? "Then I give
- 32:06the result to Claude, who for some
- 32:08reason is always arrogant, and I write
- 32:10to it:" Listen, by the way, here is the
- 32:12answer from DeepSeek and Gemini: "What
- 32:14do you think?
- 32:15The model will find the weak points.
- 32:18Then I go back to Gemini and say:"
- 32:20Claude thinks you're stupid.
- 32:22"Here are its arguments. Do you have
- 32:24anything to say in response? Right. It's
- 32:26interesting because now I am smarter.
- 32:29Because the simplest and most common use
- 32:31is: I have a question, I ask it and get
- 32:34an answer, I get lazy, copy it somewhere
- 32:36, and that's it.
- 32:38Yes, a waste of intelligence. That's
- 32:43point three.
- 32:44Point four, I don't know how else to put
- 32:47it, but in case of an unforeseen event,
- 32:50I would ask people to master what I call
- 32:53analog skills.
- 32:56Learn to grow food, bake bread, cook.
- 32:59Learn to identify edible mushrooms.
- 33:03Learn to make furniture, fix cars. Get
- 33:08it? Such knowledge in times of such high
- 33:10uncertainty could very well prove to be
- 33:13very useful.
- 33:14Even if, you know, we're not talking
- 33:16about a disaster, an AI getting out of
- 33:19control and threatening humanity's
- 33:20existence.
- 33:23But if the economy becomes so radically
- 33:26stratified between Elon Musk with a
- 33:28trillion-dollar fortune and hundreds of
- 33:30millions of starving people in the U.S.
- 33:33, or, you know, at least those who can't
- 33:35make ends meet, then these skills will
- 33:38become a shared resource that will help
- 33:40the entire community.
- 33:43The fifth skill I can call the most
- 33:45important of all.
- 33:47Interpersonal connections. And it's
- 33:49curious that in a world where machines
- 33:52can do many of the things our brains can
- 33:55do, what our hearts, our empathy, and
- 33:58our ability to understand and feel each
- 34:01other can offer becomes increasingly
- 34:04valuable.
- 34:06Look, it's easy to see that companies
- 34:09that replaced all their call center
- 34:12operators with AI in 2025—there was an
- 34:16article then about how most of them lost
- 34:19a lot of money, primarily on computing,
- 34:23but also on errors and customer churn.
- 34:26So maybe AI could help support staff
- 34:30provide people with soulful, humane
- 34:33service. And finally, point six is,
- 34:36likely, the skill that will save our
- 34:38world. Ethics.
- 34:40The problem of our time is that AI
- 34:43reflects what humanity is today.
- 34:47Which is logical, at least until
- 34:49machines become fully autonomous and
- 34:51independent. They will amplify what
- 34:55humans are. Do you understand? And we
- 34:58just suck. We live in lawlessness.
- 35:02Without the rule of law, people get away
- 35:05with serious crimes without question.
- 35:08Everyone is busy enriching themselves at
- 35:10the expense of others. Concepts are
- 35:13being misrepresented everywhere.
- 35:15A huge number of people are struggling
- 35:18to survive, dying, or losing their
- 35:21well-being and livelihoods.
- 35:24And all this for the benefit of a
- 35:26minority.
- 35:28Just imagine what would happen if such a
- 35:31world were powered up; it would become
- 35:33uninhabitable.
- 35:36Therefore, for us to cultivate a vision
- 35:39of humanity that truly works for
- 35:42everyone, we must return to ethics.
- 35:46We must return to a world where, if
- 35:48someone takes a child to an island to
- 35:51harm them, the guilty party faces the
- 35:54law, so that this doesn't become our
- 35:56reality.
- 35:58>> Thank you very much. Let me briefly
- 36:00summarize.
- 36:01First, don't lose touch with reality and
- 36:05rely on facts. Second, use AI to empower
- 36:08yourself.
- 36:10That is, if we have an IQ of 105, with
- 36:12AI we might be able to operate at a
- 36:15level of around 500.
- 36:17The point is to grow stronger thanks to
- 36:20AI, not weaker, and thus preserve your
- 36:22value.
- 36:23Third, verify incoming information,
- 36:26compare several answers, use different
- 36:29AIs, and try to find the truth yourself
- 36:32.
- 36:33Fourth, develop analog skills, such as
- 36:36singing and dancing.
- 36:39This is what AI essentially cannot do.
- 36:42Although I am curious. Do you think AI
- 36:44will eventually be able to master that
- 36:47too?
- 36:48Fifth, build interpersonal connections
- 36:51and communication. And sixth, ethics.
- 36:56Did I remember that correctly?
- 36:57>> Correct.
- 36:58>> Yes.
- 36:59>> Can we name the professions that AI will
- 37:02replace first? What do you think awaits
- 37:05us in the near future?
- 37:08And are there specialties that will last
- 37:11another 10-15 years, or will AI
- 37:13ultimately push people out of everything
- 37:16?
- 37:17>> So, the assumption that AI could replace
- 37:20all specialties is absolutely valid.
- 37:23Yes, the economic reality of our world
- 37:26makes this theory impossible.
- 37:29So, let me explain that. I will try to
- 37:33look at the question from two sides.
- 37:36On one hand, if people no longer work
- 37:38and have no income, one of two things
- 37:41could happen: either the population
- 37:43revolts and civil unrest begins, or
- 37:45governments decide to eliminate everyone
- 37:48.
- 37:49And that is not the most favorable
- 37:51outcome.
- 37:53Or the state will pay what is usually
- 37:55called universal basic income, so that
- 37:58people can lead a decent life while they
- 38:00are unable to work.
- 38:02A very controversial solution. But the
- 38:05authorities haven't come up with
- 38:07anything else yet. The problem with UBI
- 38:09is that the math just doesn't add up,
- 38:12right?
- 38:13Because if you take a country like the
- 38:17United States, where if you pay every
- 38:21adult over 18, say, $ 16,000. That is a
- 38:26fraction of today's minimum wage, then
- 38:29you would have to transfer $ 5 trillion
- 38:33a year to American citizens.
- 38:37You see, with a sum like that, the
- 38:39government needs to find these funds
- 38:41annually, considering that U.S. budget
- 38:44revenues, excluding income tax paid by
- 38:46working people, are about 1.5 trillion.
- 38:51Right?
- 38:52So, 1.5 trillion comes into the budget,
- 38:54but five needs to be paid.
- 38:57And so, from a game theory perspective,
- 39:00the only option for the state is to tax
- 39:03companies using AI more heavily.
- 39:08But since such corporations currently
- 39:09have great influence, and since people
- 39:11no longer work for them, it will be very
- 39:13difficult for the government to achieve
- 39:15this.
- 39:17So it's easy to imagine that a situation
- 39:19where no one is working and people stay
- 39:21home will either lead to civil unrest,
- 39:23or, you know, they will have to pay,
- 39:25which is impossible.
- 39:27That is one side, the governmental one.
- 39:30The corporate side, in turn, is
- 39:32languishing from human greed.
- 39:35Why? Because today, any CEO thinks:" Oh
- 39:39my God!
- 39:40"" AI can get rid of some of my staff,
- 39:42cut costs, and increase productivity.
- 39:44"Great news. I'll fire half the company
- 39:47.
- 39:48The catch is that we all understand that
- 39:52measuring GDP and economic activity
- 39:55relies mainly on consumption.
- 39:58In American GDP, 71%is consumer demand.
- 40:03If people don't have money to buy, then
- 40:05companies have no reason to produce, and
- 40:08the economy collapses.
- 40:10So, another reason for the government to
- 40:14say:" You know what?
- 40:16You can't fire people. You are simply
- 40:19required to keep them. "That is exactly
- 40:21what the authorities in China did.
- 40:23They stated outright:" Listen, you are
- 40:25forbidden from replacing anyone with AI
- 40:27; why are you offloading the problem
- 40:28onto us?
- 40:30"" Fine, we are the People's Republic of
- 40:32China, we want the whole nation to have
- 40:34a decent standard of living.
- 40:36"And in practice, that means you keep
- 40:41them. You are a profitable corporation,
- 40:43you pay them.
- 40:45And by the way, you train them so they
- 40:48can use AI so well that they become
- 40:50smarter than they were.
- 40:53Together with AI, they form a hybrid
- 40:55team that is better than a human or AI.
- 40:57Individually. Right?
- 41:00So this idea could spawn new productive
- 41:03approaches for how economies use AI to
- 41:06become smarter.
- 41:09Ultimately, people will continue to work
- 41:1210 hours a week, for example.
- 41:15But they will work, produce, develop,
- 41:17and build a better world for all of us.
- 41:20This is a viable solution. Under other
- 41:23strictly capitalist options, the math
- 41:25just doesn't add up for them. Now, the
- 41:27other side, in my opinion, is more
- 41:29interesting if you think about it.
- 41:32It's about the existence of an
- 41:34ideological challenge to the very idea
- 41:37of work around the world.
- 41:39Right? Take the West, for example. The
- 41:42Western world says:" I don't care. ""
- 41:44Let me get rich, and let everyone else
- 41:47starve. "That's normal. Well, they are
- 41:49the ones who will have to struggle with
- 41:51this choice.
- 41:53Either civil unrest or UBI.
- 41:55And and none of this is working.
- 41:58Other countries in the Global South are
- 42:00structured quite differently.
- 42:02The People's Republic of China cares
- 42:04about its people.
- 42:06So, Beijing will find ways to ensure
- 42:08citizens keep living without the
- 42:10government having to pay for them.
- 42:13Ideologically, this is closer to
- 42:14socialism or communism, which is
- 42:15acceptable for the Chinese. Right?
- 42:18You see, in Global South countries—for
- 42:21example, in my homeland where I was born
- 42:23, Egypt—you know, most of the economy
- 42:26isn't automated, so there's no talk of
- 42:28AI replacement.
- 42:31It is remarkable to me that it will take
- 42:3510, 15, or 25 years before a robot
- 42:37replaces an auto mechanic in Egypt.
- 42:42And likely, even then, people with old
- 42:44cars they can't replace due to economic
- 42:47conditions will still go to their cousin
- 42:50the mechanic, and there will be a barter
- 42:53economy.
- 42:55But all these various solutions have a
- 42:59common trait. No government is thinking
- 43:02about them. Okay? Except, I think, for
- 43:05the Chinese leadership.
- 43:07But everyone else is just pushing the
- 43:10problem off until the storm breaks. And
- 43:12that will happen soon.
- 43:14Interestingly, not because of AI, but
- 43:16ironically, because CEOs will use AI as
- 43:19an excuse for layoffs during an economic
- 43:21downturn.
- 43:23Yes, in the event of a financial crisis
- 43:26, which many signs suggest will happen
- 43:28soon, top managers won't say," I'm
- 43:30firing people because the company is in
- 43:32trouble.
- 43:34"The stock price would drop then.
- 43:36Instead, they will say," I'm cutting
- 43:37staff because I can replace the
- 43:39employees. "And rightly so. Neural
- 43:41networks are being used as a pretext,
- 43:43not for productivity growth.
- 43:44Unfortunately, we cannot count on every
- 43:47country introducing a universal basic
- 43:49income.
- 43:51That would be a positive scenario where
- 43:53people might only have to work, for
- 43:55example, 10 hours a week.
- 43:58But let's discuss another perspective,
- 43:59where companies start replacing
- 44:01employees with artificial intelligence.
- 44:03Who will these changes affect first?
- 44:07Max Tegmark was in my documentary. I
- 44:10adore him. He is incredibly smart.
- 44:12Max laughed loudly, saying essentially
- 44:16that, you know, most CEOs are confident
- 44:19they can replace everyone in their
- 44:22company. What they don't understand is
- 44:25that AGI will replace humans in
- 44:27everything, including the role of the
- 44:29boss. They themselves will get fired too
- 44:32.
- 44:33It's easy to imagine that all jobs
- 44:35requiring typing on a computer will
- 44:38disappear first.
- 44:40So, effectively, all intellectual work
- 44:43can easily be handed over to machines.
- 44:46First the most routine part, then
- 44:48something a bit more meaningful, then
- 44:51more subtle and complex work.
- 44:53So, you know, everything you used to do
- 44:57yourself, like booking travel tickets,
- 45:01AI will definitely take over. Graphic
- 45:04designers will be replaced too—that is,
- 45:07all computer-related occupations. Next
- 45:09in line are researchers or paralegals,
- 45:12financial analysts, business analysts,
- 45:15and so on. The time will come when these
- 45:18roles will also be automated—
- 45:20professions that require discipline. For
- 45:23example, project managers, they will
- 45:25likely be
- 45:26>> what is the timeline for their dismissal
- 45:28,
- 45:28>> when?
- 45:29>> Yes,
- 45:31>> I think we will see a gradual
- 45:33replacement in 2027, and then it will
- 45:35likely accelerate if the economy
- 45:38collapses in
- 45:40>> the next year.
- 45:42>> Yes, I mean, you need to understand that
- 45:45the data hides one very interesting
- 45:48aspect. Yes, the US labor market—again,
- 45:51that’s where most of my statistics come
- 45:53from, but I think it’s relevant for
- 45:55everyone.
- 45:57The American labor market only began to
- 45:59surprise on the downside in the last 3
- 46:01months, but this has impacted hiring.
- 46:05So, while layoffs weren't that frequent
- 46:08, many weren't hired, and most young
- 46:12graduates or entry-level employees
- 46:14aren't getting as many opportunities as
- 46:18they used to, because their positions
- 46:21are so entry-level that they are easily
- 46:24automated, and over time, I think AI
- 46:28will replace all intellectual
- 46:30professions and strategic roles.
- 46:35you know, managerial and so on, but that
- 46:39could take 5 to 10 years.
- 46:44What’s interesting here, again, in my
- 46:46opinion, is that if you lose your job,
- 46:48finding another one will be very, very
- 46:51difficult. Yes, intellectual ones. And
- 46:53only then will it be the turn of
- 46:56blue-collar professions.
- 46:58By that time, robots will already be
- 47:00capable of taking on some actual
- 47:02physical labor.
- 47:03All of this is just a probability. All
- 47:06right. The reality is that something
- 47:09like this would crash the global economy
- 47:12.
- 47:13And I keep repeating to the politicians
- 47:16I know that if they allow this to happen
- 47:19, their economy and their government
- 47:22will disappear before the jobs do.
- 47:26The reason is quite curious.
- 47:29You don't need to lose 100%of jobs for
- 47:31the system to collapse.
- 47:33The economy will stall at just 10-15%
- 47:36unemployment.
- 47:39>> Yes, of course, whenever we talk about
- 47:42the future, we are talking about
- 47:44probabilities.
- 47:45>> I started this podcast when I was
- 47:47studying at Moscow State University and
- 47:50knew that after graduating I would find
- 47:53a job.
- 47:54And I think many students, listening to
- 47:57our episode, are wondering, maybe we
- 48:00should drop out of university or, maybe
- 48:03I am advising them to do just that.
- 48:06100%.
- 48:07>> Even Harvard.
- 48:08>> 100%.
- 48:09I was sitting with a wonderful girl, the
- 48:12daughter of a close friend, and she is
- 48:14just about to enter Harvard next year.
- 48:17I said:" Kni, by the time you graduate
- 48:19from university in 4 years, the world
- 48:20will be completely different.
- 48:22"Really?
- 48:23Universities never keep up, do they? Of
- 48:26course, there is huge value in
- 48:28interacting with smart people. It helps
- 48:32to master life skills.
- 48:36But knowledge, it's a pity, but the
- 48:38storage of knowledge today is entirely
- 48:41handed over to machines. It has been
- 48:43transferred to computers.
- 48:45And there are skills that children or
- 48:47young people can acquire to become
- 48:49smarter without going to university.
- 48:52Now I will say something very, very
- 48:54directly. So, Emma is my startup. If I
- 48:57had started this business in 2022, no
- 49:00joke. I would have needed 400 engineers
- 49:03and probably four years of work.
- 49:07This is my 39th startup, so I know what
- 49:09I'm talking about.
- 49:11So, it took 9 months to create Emma, and
- 49:14there are two of us.
- 49:16And it's just amazing that this is
- 49:19possible today.
- 49:21So, for every young graduate or anyone
- 49:24entering the job market, I want to open
- 49:27their eyes to a different approach.
- 49:30Look, you used to strive to graduate
- 49:33from university to get what? A diploma.
- 49:37That's exactly what you were buying.
- 49:40They sold you a piece of paper that said
- 49:42," He was with us for four "or" She was
- 49:44with us for 4 years.
- 49:46"That's fine, even more than that, it's
- 49:49great. Okay.
- 49:51And with that diploma, you would go to a
- 49:53company and say," Could you please hire
- 49:55me, pay me, and enslave me for the rest
- 49:56of my life?
- 49:58"This paradigm is changing in two very
- 50:00interesting ways.
- 50:02Companies won't be hiring as much, but
- 50:04what's even more curious is that they
- 50:06can be truly disrupted.
- 50:09And I think the biggest opportunity
- 50:10offered by AI is what I call the
- 50:12replacement cycle.
- 50:14Let me give you a very simple example. I
- 50:15am old enough to remember how I used to
- 50:17listen to music. My first recordings
- 50:21were on vinyl, and I had quite a few.
- 50:25There were bands I absolutely adored.
- 50:27And then they said," No, no, no, now
- 50:29it's cassettes, they're more convenient
- 50:31. You can take them in the car and so on
- 50:33. "And I replaced all my vinyl with them
- 50:35.
- 50:36Then CDs, there were Walkmans, you could
- 50:39carry them with you. Discs arrived with
- 50:41much better sound.
- 50:44I threw away all the cassettes and
- 50:45bought a bunch of discs.
- 50:48Every time new technology appeared, I
- 50:50changed my whole collection for a new
- 50:51one.
- 50:52Right.
- 50:54And then Apple tricked me, saying," No,
- 50:56no, you don't need discs anymore. "I
- 50:59threw them out. Screw them. Right. In
- 51:01the end, I didn't own my music anymore.
- 51:04I had to rent it from Apple. Eventually
- 51:07, I went back to vinyl. So, there is a
- 51:09quite interesting thought here. This
- 51:11cycle will repeat itself. Think about it
- 51:14.
- 51:14There was a time when you had to buy
- 51:17Microsoft Office and pay, I don't know,
- 51:19a lot of money per year.
- 51:22>> You see? I'm not joking. Leaving this
- 51:25studio, even before I walk to my car, I
- 51:28can write a prompt in Claude Code, and
- 51:31by the time I reach my destination, I
- 51:33will have my own Microsoft Office.
- 51:37Okay?
- 51:38Such a replacement cycle is inevitable.
- 51:41This is, essentially, understand, the
- 51:44very idea that I have to sit down and
- 51:48Making a presentation is foolish.
- 51:52The approach where I have to type to
- 51:55create a document is not user-oriented.
- 51:59Otherwise, I would just talk to a neural
- 52:01network as if it were a junior employee
- 52:04. And we discuss what I want to write
- 52:06and what my opinion is on the subject.
- 52:09What design to choose and so on, and
- 52:11then I would just ask it to prepare the
- 52:13text for me.
- 52:16And all this opens up huge scope for two
- 52:18things.
- 52:20First: reinventing all technologies in a
- 52:23new replacement cycle.
- 52:25But far more interesting is the chance
- 52:27to do the same on a sovereign basis.
- 52:29What do I mean?
- 52:32If you take a country like the UK—and
- 52:34I’m just speculating here, so don't
- 52:35quote me.
- 52:37They spend somewhere between 7 and 10
- 52:39billion dollars on Microsoft Office,
- 52:41which leaves the country for the US
- 52:43annually.
- 52:45If the British economy wants to save
- 52:48itself in a developing world, it can
- 52:51simply apply a fiscal policy that says:
- 52:55" We have a new product, a British
- 52:57equivalent of Office.
- 53:00It is available to every citizen and
- 53:03every company in the UK for free. "Okay
- 53:05.
- 53:06By the way, the development costs
- 53:08practically nothing.
- 53:09>> This would change the world
- 53:11>> completely. Now think about that and
- 53:14countries like Russia, which could do
- 53:17this in a snap.
- 53:19Remember France, for example, which,
- 53:22after Donald Trump began opposing Europe
- 53:24, started saying:" No, our agencies will
- 53:27never use American products again
- 53:29because, firstly, they are expensive,
- 53:32and secondly, there is a risk that in
- 53:34the event of a disagreement, we won't
- 53:37even be able to access our email.
- 53:40"Right?
- 53:42And when the world starts to look like
- 53:44this, this replacement cycle opens up a
- 53:47mass of opportunities.
- 53:50Again, I’m not trying to promote an app
- 53:52here, but Emma, in my opinion, started
- 53:55with the question: if you were creating
- 53:57a dating app in, well, you know, the
- 53:59world, what would it look like?
- 54:02Okay, what are the real problems with
- 54:04such services? Why didn't they work? Did
- 54:06they meet the user's needs and so on?
- 54:08And suddenly I say:" If it works, we
- 54:10succeeded.
- 54:11"We created a service that works
- 54:13perfectly, knows how to introduce you to
- 54:15the right people, and doesn't force you
- 54:17to swipe through photos.
- 54:19See?
- 54:21And when we launched it, we suddenly
- 54:23realized:" Wait a minute, if we know the
- 54:26user this well, we can match them with
- 54:28friends, suggest areas of interest, and
- 54:31even establish professional connections
- 54:33.
- 54:34"With this alone, you disrupt the entire
- 54:37dating industry. A terrible market with
- 54:39an annual turnover of 90 billion.
- 54:42Everyone who looks like LinkedIn or
- 54:44looks like Meetup is at risk. The list
- 54:46is quite long.
- 54:48The idea I’m trying to get across to
- 54:51every young person is this: instead of
- 54:53wasting 4 years of your life, and
- 54:55however much your university degree
- 54:57costs, put all that into trying to get
- 55:00ahead of the curve and create a
- 55:02breakthrough product that replaces
- 55:04current technologies.
- 55:07Interestingly, the government could hire
- 55:08you and say," This is our sovereign
- 55:09office.
- 55:11"For example, our sovereign Uber. How
- 55:13many countries could you launch such an
- 55:15app in right now?
- 55:18Give it to the users, the drivers, and
- 55:19say," Hey, by the way, keep it for
- 55:21yourselves, just keep it.
- 55:23"" If you get an order through me, it
- 55:26will be 10-15%more profitable for you. "
- 55:29Right?
- 55:30>> Yes.
- 55:32>> There are so many opportunities right
- 55:34now, and this has actually captured
- 55:36everyone's attention thanks to Anthropic
- 55:38itself.
- 55:40There was a stock market crash, I think
- 55:43around April, when Dario essentially
- 55:45came out and said," Hey, something is
- 55:47going to happen to SaaS, Software as a
- 55:50Service.
- 55:51"That means every SaaS service will be
- 55:53completely rewritten. And it's true.
- 55:56Take a country like India, full of smart
- 55:58technical specialists who will rewrite
- 56:01all the code.
- 56:03So, that's just the first step.
- 56:07The replacement cycle is the beginning.
- 56:10Step two, in my view, is that the
- 56:12economy will return to something more
- 56:14similar to old agricultural models,
- 56:16where there is no global business.
- 56:19So more local small businesses will
- 56:21emerge, where everyone helps each other
- 56:23.
- 56:25Remember that the biggest lie, the lie
- 56:28of capitalism, is that you have to work
- 56:32very hard to earn a lot of money, and
- 56:35then give a large chunk away in taxes.
- 56:40Another significant share for rent to
- 56:42live closer to the office, and even more
- 56:44for office attire and transportation
- 56:46costs.
- 56:48And in the end, just to survive. Right.
- 56:51And so many people live in this paradigm
- 56:53.
- 56:54My question is: how reliable is this if
- 56:57some of these people lose their jobs?
- 57:00What will they do? They’ll probably say
- 57:02," You know what?
- 57:04I’ll get some chickens, they’ll lay eggs
- 57:06, you’ll keep bees and get some honey,
- 57:09and we’ll build small economies in small
- 57:11communities where you don't need to
- 57:14trade your iPhone for the eighteenth
- 57:16model, because in truth, it was exactly
- 57:18the same as the eleventh.
- 57:21"Since my loved ones live nearby, I
- 57:23don't need that anymore.
- 57:25These highly disruptive forces present a
- 57:28threat, but also huge opportunities to
- 57:30start living differently.
- 57:33And I think the young generation
- 57:35graduating today is asking these
- 57:38questions and saying," Things are
- 57:40getting more complicated and bringing
- 57:42less satisfaction.
- 57:44"Maybe we shouldn't participate in all
- 57:46of this anymore?"
- 57:47Yes, after this podcast, I will
- 57:49definitely master v0-coding and add it
- 57:52to my schedule, because I think without
- 57:55this skill, it will be almost impossible
- 57:57to keep up. "How much time do you
- 58:00personally spend now to stay up to date
- 58:02with what's happening in the industry? I
- 58:04spend a lot of time, probably 3-4 hours
- 58:07a day, just to stay informed. But I need
- 58:10to understand the underlying essence of
- 58:12technologies, not just their events.
- 58:16At my age, I mean, I still understand
- 58:18everything, but, you know, I haven't
- 58:21been 27 for a long time, so I really
- 58:23have to dig deep into technical
- 58:25documentation, various forums, and try
- 58:28everything with my own hands.
- 58:31That takes time.
- 58:33I recommend that everyone, regardless of
- 58:36their education or field of work, spends
- 58:38, say, an hour a week following the news
- 58:40.
- 58:41And really, it’s just a matter of a few
- 58:44searches on YouTube, watching a couple
- 58:47of videos, and liking the ones you found
- 58:50useful.
- 58:51The platform will very quickly start
- 58:53keeping you updated on everything that's
- 58:55going on. You know, your feed itself,
- 58:58your algorithm, will show you what you
- 59:00need to know. But I think the skills
- 59:03people should brush up on are not just
- 59:05about technology. I mean, you don't
- 59:07necessarily need to master coding.
- 59:11You need to learn to imagine a world
- 59:14that is AI-oriented and find your niche
- 59:17in it.
- 59:19And a world built on AI is a space where
- 59:23you have the opportunity to dream.
- 59:28If I were to redesign travel, how would
- 59:31I do it today?
- 59:35And there will always be a technical
- 59:37specialist you can team up with who can
- 59:38help you do it.
- 59:40If you decided to restructure healthcare
- 59:44, for example, there are a ton of
- 59:47possibilities.
- 59:49I think there are two skills that people
- 59:51should pay attention to. The first is
- 59:53entrepreneurship, which is the
- 59:55willingness to try and fail, try and
- 59:58fail, try and fail. I rewrote the code
- 1:00:00for Ema, or rather, the whole team did.
- 1:00:02We rewrote it nine times. Not updated it
- 1:00:05nine times, but deleted the code and
- 1:00:07started over nine times. Okay? Why not?
- 1:00:11It would have taken four years before,
- 1:00:14but now it takes a month and a half.
- 1:00:16The point is, I don't actually need a
- 1:00:18job to do something amazing.
- 1:00:21You don't need a job to do something
- 1:00:23amazing. Find your niche and develop in
- 1:00:25it. This is the new world we are going
- 1:00:28to live in. And please, stop this
- 1:00:30madness." Oh, I'm going to be an
- 1:00:32influencer. "Those days are gone, you
- 1:00:35know? Simply because there will be more
- 1:00:38AI influencers than anyone can count.
- 1:00:40And yes, you might say," No, no, no,
- 1:00:42people won't identify with them. "
- 1:00:45Believe me.
- 1:00:46You know, friends sometimes send me
- 1:00:48photos and write:" Oh my God, have you
- 1:00:50seen this girl on Instagram?
- 1:00:52Amazing.
- 1:00:53"And yes, I very often share videos of
- 1:00:56beautiful nature, and my daughter says:
- 1:00:59" No, no, Dad, that's AI. "
- 1:01:03That's one side.
- 1:01:05And the other side, look at your YouTube
- 1:01:08feed and count how many videos you
- 1:01:11watched today were voiced by a script
- 1:01:14reader and had AI-generated visuals.
- 1:01:18And it will take you a minute, or maybe
- 1:01:20even 10, to do that. And you'll realize
- 1:01:23how many videos were not made by a human
- 1:01:25. So I think that now we can live in a
- 1:01:28world where we create meaningful things
- 1:01:30because we've gotten smarter, and we can
- 1:01:33be entrepreneurs because we've become
- 1:01:35more flexible, and creating has become
- 1:01:37much easier.
- 1:01:38We are no longer forced to be employees
- 1:01:41under someone who says:" Someday I will
- 1:01:43replace you. "
- 1:01:44>> Honestly, it sounds depressing.
- 1:01:48I'm an influencer, after all.
- 1:01:49>> I think you'll be just fine. Yes.
- 1:01:52>> 100%.
- 1:01:53Because remember, the fifth skill is
- 1:01:55interpersonal connections. Yes. To be
- 1:01:57honest, there was a moment in my career
- 1:02:00when I decided, you know what? In 3
- 1:02:02years, AI will write better than me.
- 1:02:05Why am I even writing books?
- 1:02:08And what's interesting is that if I talk
- 1:02:10about my daughter and AI talks about a
- 1:02:12fictional daughter that doesn't exist,
- 1:02:14the audience is put off by the story.
- 1:02:16You, but my story really resonates with
- 1:02:19them. We still live in a world where
- 1:02:21people need people. Right?
- 1:02:24And you know, in your job as a host, you
- 1:02:27embody the curiosity of your audience.
- 1:02:30Well yes, I mean, let me give you a very
- 1:02:32simple example. Seriously, if you had
- 1:02:34sent me these questions in advance, I
- 1:02:37would have given them to Gemini. It
- 1:02:40would have written a script of how Mo
- 1:02:43Gawdat would answer, and the result
- 1:02:45would be 70%ready.
- 1:02:48I would have uploaded the text into
- 1:02:50HeyGen and sent you an avatar saying it
- 1:02:52.
- 1:02:53Would anyone watch that?
- 1:02:55You see, it is precisely in this human
- 1:02:58desire to deal with humans that the real
- 1:03:00opportunity lies.
- 1:03:03At the same time, you know, as I said, I
- 1:03:05think you need to change with the times
- 1:03:07.
- 1:03:07Maybe even become more human. Right?
- 1:03:10Maybe interact more with people. That is
- 1:03:13definitely a reliable approach. There is
- 1:03:15one more question that personally
- 1:03:17concerns me.
- 1:03:19Today, if a person is talented, willing
- 1:03:22to work hard, and has an entrepreneurial
- 1:03:25spirit, they can build a company and
- 1:03:27succeed, even if they grew up in a poor
- 1:03:30family.
- 1:03:32But what will happen if in the future
- 1:03:34the most powerful AI systems end up
- 1:03:36under the control of only a small group
- 1:03:39of people?
- 1:03:41A child born into the family of a
- 1:03:43founder or a company can get access to
- 1:03:46almost everything, while someone from a
- 1:03:49poor family will not have such
- 1:03:51opportunities.
- 1:03:53Could AI actually hinder social mobility
- 1:03:56and increase the gap between people?
- 1:03:59What do you think about this?
- 1:04:02I think there is reason to believe that
- 1:04:05this could be our near future.
- 1:04:09Overall, if you look at the trend, when
- 1:04:13you graduated in my time, in my
- 1:04:15generation, you really had a chance to
- 1:04:18work hard and buy a house and start a
- 1:04:21family and so on and so forth;
- 1:04:24graduating now, in most countries of the
- 1:04:27world, that is almost impossible.
- 1:04:32Once again, I will always come back to
- 1:04:35the fact that it is not about the idea,
- 1:04:37but about the capitalist mindset, which
- 1:04:40states that it is perfectly normal for 1
- 1:04:43%of the population to appropriate 90%of
- 1:04:45the wealth.
- 1:04:47And also the political logic, according
- 1:04:50to which, if you have money, you can
- 1:04:52influence government policy so that it
- 1:04:54favors whoever is richer.
- 1:04:57It seems to me that, unfortunately, many
- 1:05:00changes come through pain, and the world
- 1:05:04will soon have to go through it.
- 1:05:07The world is about to say:" Enough,
- 1:05:11enough wars, enough empty promises,
- 1:05:14enough lies.
- 1:05:17"And obviously, the reaction to such a
- 1:05:20thing can be even more painful, because
- 1:05:23I think those who sit in high positions
- 1:05:26will stubbornly resist.
- 1:05:30They will start relying on surveillance
- 1:05:32or control, or anything else to try.
- 1:05:35So things are getting very complicated
- 1:05:38here.
- 1:05:39Moreover, it is all because you are
- 1:05:43playing a game that favors those at the
- 1:05:47top.
- 1:05:49I think there is a sea of games that are
- 1:05:52a bit further from the city, a bit
- 1:05:54further from the frantic economy, where
- 1:05:57you won't be forced to constantly chase
- 1:06:00trends, for example, and all these
- 1:06:02opportunities will remain.
- 1:06:05And there will likely be more of them.
- 1:06:07Okay?
- 1:06:09And I am inclined to believe that those
- 1:06:12who remain under this pressure, deprived
- 1:06:15of the chance to succeed, will most
- 1:06:18likely push harder and harder and harder
- 1:06:21until something changes or breaks.
- 1:06:25In my opinion, the point is not to sit
- 1:06:27idly by and say:" The government is
- 1:06:29doing nothing, the oligarchs, the tech
- 1:06:32moguls are taking the bread out of our
- 1:06:34mouths.
- 1:06:35"But to tell yourself:" No, my piece. "
- 1:06:38They won't take it, just like they won't
- 1:06:39take the piece of my loved ones.
- 1:06:42I will do something about it, I will
- 1:06:44succeed.
- 1:06:45What if in the near future you have to
- 1:06:47have a chip implanted in your brain to
- 1:06:49use artificial intelligence more
- 1:06:50effectively?
- 1:06:51Would you go for it?
- 1:06:53And what do you think about such a
- 1:06:55prospect?
- 1:06:55>> No way.
- 1:06:58I mean, yes, it will happen, but I'm not
- 1:07:00participating.
- 1:07:01>> You think it will happen?
- 1:07:04>> Yes, I think it will happen in one of
- 1:07:06two ways.
- 1:07:07The first option: the technology will
- 1:07:10either allow a very few to become
- 1:07:12super-intelligent compared to the rest.
- 1:07:16It is not hard to imagine that such a
- 1:07:18chip will be available only to the rich
- 1:07:21and famous, if you will, to the
- 1:07:22detriment of us all.
- 1:07:25Which, by the way, is typical for
- 1:07:27technology.
- 1:07:28You know, people who bought the first
- 1:07:30iPhone paid much more for it than, for
- 1:07:32example, when the technology became
- 1:07:34affordable thanks to Android.
- 1:07:36That's the first option. Another
- 1:07:39possibility is that governments will use
- 1:07:42the chips for control. But in any case,
- 1:07:44I would prefer not to participate.
- 1:07:47I have seen a lot and lived a long
- 1:07:49enough life to love the messiness of my
- 1:07:52own brain and life, and to love holding
- 1:07:54my sweetheart's hand.
- 1:07:56I love this life. After all, there are
- 1:07:59many new trends around right now. For
- 1:08:02example, human enhancement with
- 1:08:04intelligence.
- 1:08:07Creating hybrid humans, essentially part
- 1:08:11robots.
- 1:08:13There is a lot of talk about longevity,
- 1:08:16about ways to live to be 100 or even 500
- 1:08:19years old.
- 1:08:20None of this is for me. I
- 1:08:25I love being human, and I will remain
- 1:08:27one, hopefully smarter thanks to any
- 1:08:30tools that will be available to me in
- 1:08:32the next couple of years.
- 1:08:36And when the time comes to stop
- 1:08:38competing, I will go fix cars.
- 1:08:42But do I understand you correctly?
- 1:08:46If you don't use brain-implanted chips,
- 1:08:48you'll fall behind those who do, because
- 1:08:51those people will be smarter, become
- 1:08:53better entrepreneurs, and you won't be
- 1:08:55able to compete with them.
- 1:08:58How are we supposed to deal with that?
- 1:09:00It all depends on whether you decide to
- 1:09:02compete.
- 1:09:04I have already said that two worlds will
- 1:09:06emerge.
- 1:09:07One world is incredible, but they will
- 1:09:09concentrate all the resources.
- 1:09:11Unless you have a tiny patch of land and
- 1:09:15a couple of chickens.
- 1:09:18Of course, life, in general, I must put
- 1:09:20it bluntly.
- 1:09:23If you have ever lived in a surfer
- 1:09:26village or, you know, in the Caribbean
- 1:09:30or a small village in India, or let's
- 1:09:34say, if you went hiking somewhere in
- 1:09:37Nepal, you would realize that life is
- 1:09:41not at all what you are told. So we, as
- 1:09:45humans, need something very, very
- 1:09:47different. And your purpose as a human
- 1:09:49is also quite different.
- 1:09:52Somewhere along the way, we were tricked
- 1:09:54into believing that work is the meaning
- 1:09:57of life and wealth is the only goal.
- 1:10:00Right?
- 1:10:02I don't think it's really like that.
- 1:10:04Actually, I know we are talking about
- 1:10:06technology, but allow me to philosophize
- 1:10:08a little.
- 1:10:09>> Of course,
- 1:10:10>> if you want to understand the essence of
- 1:10:13humanity's purpose, go back to the first
- 1:10:16humans, who had only one goal: to live.
- 1:10:21You see, you know, in those days there
- 1:10:23were no Ferraris, no larger apartments,
- 1:10:26no mortgages.
- 1:10:28All of this is a reaction to the
- 1:10:31simulated world we have built.
- 1:10:34So, but many people will tell you that,
- 1:10:37you know, living
- 1:10:40is much simpler.
- 1:10:44Truly enjoying life is much easier than
- 1:10:46in the world we have created.
- 1:10:50So, let's say, to give a simple example
- 1:10:53, being in shape.
- 1:10:5520 years ago, it meant going to the gym
- 1:10:58and running from time to time.
- 1:11:01Staying in shape today means you have to
- 1:11:03buy every expensive fitness tracker,
- 1:11:05constantly stare at the numbers, and
- 1:11:07count every single calorie.
- 1:11:09What nonsense, right?
- 1:11:12When did you convince me that to take
- 1:11:14care of my own body, I have to report
- 1:11:16everything to you?
- 1:11:19And I realize this sounds very strange
- 1:11:22coming from an expert, but everyone I
- 1:11:24know and everyone I have worked with
- 1:11:26wanted to use technology to make
- 1:11:28people's lives better.
- 1:11:30And we actually succeeded.
- 1:11:33You know, for a long time I managed half
- 1:11:35of Google worldwide and opened more than
- 1:11:38half of Google's offices and divisions
- 1:11:40on the planet in over 103 languages,
- 1:11:43with deep understanding of those markets
- 1:11:45, e-commerce, and launching the internet
- 1:11:47there.
- 1:11:50We as a team probably reached 4 billion
- 1:11:52people who had never been reached by the
- 1:11:53company or the internet before.
- 1:11:56We really brought about change.
- 1:11:59However, in technology, there comes a
- 1:12:01moment when promises are forgotten for
- 1:12:04the sake of capitalist goals, and value
- 1:12:06is no longer provided.
- 1:12:09Okay?
- 1:12:11In life, there comes a time when you ask
- 1:12:13yourself:" Do I need more Instagram
- 1:12:15followers, or is it better to put this
- 1:12:18away, take my girl, and meet up with my
- 1:12:20best friend?
- 1:12:23"And I sincerely believe that somehow we
- 1:12:26have all been convinced that this is not
- 1:12:29an option.
- 1:12:31It is the only option. Moreover, it is
- 1:12:34the path that most people come to as
- 1:12:37they gain wisdom. There will be two
- 1:12:39worlds.
- 1:12:40One competitive, ruthless, almost like
- 1:12:43in the movie Elysium, if you've seen it
- 1:12:46, or whatever it is, Wall-E, yes, the
- 1:12:49cartoon, where there will be a world
- 1:12:51built around wealth and AI, and the
- 1:12:54other completely, I wouldn't say
- 1:12:56primitive, but more natural.
- 1:12:59>> Yes.
- 1:13:01So, in my opinion, in my opinion, unlike
- 1:13:05what the novel predicts, if this world
- 1:13:12is not attacked by those who want more
- 1:13:15and more and more—which, unfortunately,
- 1:13:18is typical of history—a simple life
- 1:13:21doesn't mean that Europeans won't come
- 1:13:24and take over your continent and, you
- 1:13:27know, commit ethnic cleansing, and then
- 1:13:29call those lands America.
- 1:13:32Right? Alas, that is usually how it
- 1:13:35happens when people say," I don't want
- 1:13:38to play this game anymore.
- 1:13:41"However, according to my analysis,
- 1:13:43neural networks will prevent this.
- 1:13:46Yes, the calculations show that when AI
- 1:13:49becomes fully autonomous,
- 1:13:50superintelligence, by definition, will
- 1:13:53be benevolent.
- 1:13:55For the simple reason that
- 1:13:58a lot in game theory, applied
- 1:14:00mathematics, strategic analysis, quite a
- 1:14:03bit in understanding physics around the
- 1:14:05principle of least energy, as well as
- 1:14:08much in life science, especially related
- 1:14:10, you know, to evolutionary biology and
- 1:14:13the idea of the transition from kin
- 1:14:15selection to the expansion of the moral
- 1:14:17circle.
- 1:14:19A lot of scientific data indicates that
- 1:14:22the more cognitively developed a species
- 1:14:25is, the more it tends to protect life.
- 1:14:28Right? In my opinion, yes, there is a
- 1:14:31possibility that it could destroy us all
- 1:14:34.
- 1:14:35Actions, as we know,
- 1:14:37>> remember, we are not afraid of AI at all
- 1:14:40, we are afraid that it will become
- 1:14:43Donald Trump.
- 1:14:45Yes, that’s what scares us.
- 1:14:48And I think many understand that the
- 1:14:50problem is not an abundance of
- 1:14:53intelligence.
- 1:14:54Essentially, intelligence itself has no
- 1:14:56polarity.
- 1:14:57>> I heard they have a contract with the
- 1:15:00Pentagon.
- 1:15:00>> Of course, absolutely. Again, when I
- 1:15:04talked about ethics, I meant exactly
- 1:15:06what I am asking people to do.
- 1:15:10If you care about your children, do not
- 1:15:13support a business that, most likely,
- 1:15:15will end up not caring what happens to
- 1:15:17them.
- 1:15:19Okay?
- 1:15:20This is, again, part of the habit, part
- 1:15:22of the thrill.
- 1:15:24Yes, but, Mo, I have nothing else to do
- 1:15:27.
- 1:15:28I need to scroll through Instagram.
- 1:15:30No, you don't need to scroll through it
- 1:15:32. So you don't have an OpenAI
- 1:15:34subscription?
- 1:15:35Of course not.
- 1:15:37Neither do I. I only have Claude and
- 1:15:39Gemini.
- 1:15:40>> I mainly use Gemini Psychic.
- 1:15:43I have Claude too, but it seems to me
- 1:15:45that lately it has become much more
- 1:15:47arrogant than it needs to be.
- 1:15:50And that occupies me, because there is
- 1:15:53one thing in my life.
- 1:15:57Remember, among the skills I suggested
- 1:15:58we think about, number three is debating
- 1:16:00, trying to find the truth.
- 1:16:03I think one of the most important things
- 1:16:05we, as humans, need to realize. We have
- 1:16:08a choice, and a huge one at that.
- 1:16:12We can replace all technology with
- 1:16:14sovereign ones, we can put down our
- 1:16:16phones and go hug our friends.
- 1:16:19Nothing stops us from quitting our jobs
- 1:16:21and starting to live more simply. We can
- 1:16:23. There are many options.
- 1:16:27We are distracted enough not to spend
- 1:16:30time thinking and scared enough to
- 1:16:32consider such things madness.
- 1:16:36Yes, but I can easily say to you:" Go to
- 1:16:38India at least once, and you will
- 1:16:40understand how simply one can live.
- 1:16:44"Yes, you don't have to live like the
- 1:16:46local poor, but there are so many
- 1:16:48opportunities that don't require you to
- 1:16:51wear a fitness tracker, implant a chip
- 1:16:53in your head, and lead a lifestyle
- 1:16:55increasingly resembling the life of a
- 1:16:58cyborg.
- 1:17:00That's a very Californian view of
- 1:17:02existence. But we are not all
- 1:17:04Californian.
- 1:17:05>> We are just humans. Yes,
- 1:17:10I’m afraid I might be wrong, but I
- 1:17:12thought Elon Musk and Sam Altman were in
- 1:17:15your documentary.
- 1:17:17Did I see them?
- 1:17:18>> No, they weren't.
- 1:17:19>> But did you include them?
- 1:17:21>> We made a very interesting decision
- 1:17:23while working on the documentary.
- 1:17:25Namely, we didn't want to interview
- 1:17:28players who, in our opinion, are not
- 1:17:31making the world a better place.
- 1:17:35We took their speech from open access
- 1:17:38sources.
- 1:17:39>> Understood.
- 1:17:40>> Yes. To show what they really stand for
- 1:17:43, I didn't need to sit at the same table
- 1:17:46as Peter Thiel when another interviewer
- 1:17:49asked him," Would you prefer that
- 1:17:51humanity survive?
- 1:17:53"And Peter Thiel was like," Well, that's
- 1:17:55a complicated one.
- 1:17:57"No, I, you know, that's enough to
- 1:18:01understand everything.
- 1:18:04You don't need to be in the same room
- 1:18:06with them to ask such things.
- 1:18:08The question has already been asked, and
- 1:18:10the answer is publicly available. So,
- 1:18:11it’s not necessary.
- 1:18:13There was a clip in the film with Sam
- 1:18:15Altman where he basically says," Well,
- 1:18:17you know, it will likely lead to the end
- 1:18:19of humanity or its extinction.
- 1:18:22"I don't remember exactly.
- 1:18:23>> He said that back in '15,
- 1:18:24>> yes. But until then, many good companies
- 1:18:28will appear.
- 1:18:32We mostly talked with Geoffrey Hinton,
- 1:18:34Max Tegmark, Yoshua Bengio, and Fei-Fei
- 1:18:37Li.
- 1:18:38You know, with many prominent scientists
- 1:18:41who are essentially saying," This is our
- 1:18:43path, we were doing this, we were
- 1:18:45creating that. "Yes, we’ve done this
- 1:18:48before and we know where it all leads.
- 1:18:50And I think that's exactly what we need
- 1:18:51to be talking about.
- 1:18:53Elon Musk said that we are in the early
- 1:18:55stages of a technological singularity.
- 1:18:58And do you agree with him?
- 1:18:59>> 100%.
- 1:19:01I don't think we will make it past 2027
- 1:19:04without reaching the singularity.
- 1:19:06>> And what would that mean for humanity?
- 1:19:10>> It implies a rapidly accelerating path
- 1:19:13toward superintelligence.
- 1:19:18So with AI at its current level, with
- 1:19:20recursive self-improvement and long-term
- 1:19:23reasoning, it becomes crystal clear that
- 1:19:26we will reach a point where AI, either
- 1:19:28collectively or as individual systems,
- 1:19:31will become smarter than humans.
- 1:19:35And that would mean that things will
- 1:19:38happen in our world that are very
- 1:19:40difficult to predict.
- 1:19:43Anyone who tells you they know what 2028
- 1:19:46will look like is being too arrogant,
- 1:19:48right?
- 1:19:49So, the concept of a singularity in
- 1:19:51physics is an event horizon beyond which
- 1:19:53you cannot predict what will happen
- 1:19:55because the laws of nature no longer
- 1:19:57apply there.
- 1:20:00True, at the edge of a black hole, for
- 1:20:03example, the singularity is also about
- 1:20:05an event horizon where artificial
- 1:20:08superintelligence, or at least
- 1:20:10artificial general intelligence, becomes
- 1:20:13so viable that it's impossible to
- 1:20:15predict what the world will be like.
- 1:20:19>> Does this mean that as humans we won't
- 1:20:21be able to live the way we do today?
- 1:20:24>> In the future, our lives will be more
- 1:20:26human. All right. Once again. Remember,
- 1:20:29as I already explained, AI companies
- 1:20:32have an incentive to say that they will
- 1:20:34replace workers. Okay? The capitalist
- 1:20:37system finds it profitable to convince
- 1:20:40you that you cannot live without it. Now
- 1:20:42I am going to give you a very important
- 1:20:44hint.
- 1:20:47Capitalism wants you to believe that you
- 1:20:50will eventually become obsolete, while
- 1:20:54simultaneously trying to convince you to
- 1:20:57work as hard as possible because you
- 1:21:00cannot be replaced yet.
- 1:21:04Again, my entire philosophy is that you
- 1:21:06have a choice. You can decide to
- 1:21:08completely step out of this, so to speak
- 1:21:10, system and live in parallel economies
- 1:21:12.
- 1:21:13And again, if you grew up in the Western
- 1:21:15world, you might think such things don't
- 1:21:17exist.
- 1:21:18Yet it is a large part of the planet.
- 1:21:23Most of the population in the Global
- 1:21:25South lives in a way that is completely
- 1:21:27unlike California, New York, London, and
- 1:21:29so on.
- 1:21:31They are cheerful, active, cohesive, and
- 1:21:33warm societies where people sing, dance
- 1:21:36, eat the best food, and laugh from the
- 1:21:38heart every day. They simply have no
- 1:21:40need to buy a Ferrari. Yes. Yes,
- 1:21:43>> and you have that choice, but there is
- 1:21:45an even more interesting option.
- 1:21:48And it is unclear to me why humanity
- 1:21:50still hasn't taken advantage of it.
- 1:21:53If you guys are planning to replace us
- 1:21:55in 5 years, you know what?
- 1:21:57We quit now, because ironically, again,
- 1:22:00you are told that since you might be
- 1:22:02replaced soon, you should work until you
- 1:22:05drop to save a little for a rainy day.
- 1:22:09But the truth is that if you seriously
- 1:22:12declare:" No, no, wait, we are still
- 1:22:15needed for labor arbitrage.
- 1:22:18"So this is the only final moment when
- 1:22:20we have the power to state:" Here are
- 1:22:21our demands.
- 1:22:23"Okay, here is this choice. It exists
- 1:22:25regardless. An interesting challenge.
- 1:22:28And the reason one of the books I'm
- 1:22:30working on right now is called
- 1:22:31Propaganda.
- 1:22:33It's a curious problem because people
- 1:22:35are convinced they are helpless.
- 1:22:38That is not the case. We have the power
- 1:22:40to deprive malicious players of our
- 1:22:42income. You see?
- 1:22:45We have the power to support good-faith
- 1:22:47participants who advocate for well-being
- 1:22:50and use their technologies.
- 1:22:52We have the choice to forgo technology
- 1:22:54altogether. Right?
- 1:22:57We can choose not to drink soda, or if
- 1:23:00we do, we can buy it from a national
- 1:23:02company to keep wealth sovereign.
- 1:23:06Yes, or should we prefer juice? The
- 1:23:10decision is up to us. The question is
- 1:23:12for the individual, when will they stop
- 1:23:13and say:" I have options before me. I
- 1:23:15will take advantage of one of them.
- 1:23:18"And, of course, if each of us makes a
- 1:23:20choice and declares—and the conclusion
- 1:23:23of the novel is brilliant—it says:" I am
- 1:23:26not asking to stop AI.
- 1:23:28"" I am suggesting that we create narrow
- 1:23:33AI that serves humanity, "
- 1:23:37>> yes, ethical,
- 1:23:39>> yes.
- 1:23:40So when a company wants to create AI to
- 1:23:42cure cancer, it approaches the state,
- 1:23:44submits an application, and the
- 1:23:46regulator replies:" Wonderful, we
- 1:23:47support you.
- 1:23:49"Go ahead, defeat the disease.
- 1:23:51But if someone else says," I'm
- 1:23:53developing AI to hack companies or play
- 1:23:56the stock market for my own gain, "the
- 1:23:59authorities say:" that's a waste of
- 1:24:01resources. "That is selfishness.
- 1:24:04Find an AI that will serve humanity.
- 1:24:08If a country decides not to limit
- 1:24:10artificial intelligence, but allows it
- 1:24:12to develop as fast as possible and gives
- 1:24:14it military access, it could gain a huge
- 1:24:17advantage.
- 1:24:19But if another country does the same,
- 1:24:21victory might go to them instead.
- 1:24:24So we are facing a kind of dilemma.
- 1:24:28No one can afford to hold back because
- 1:24:31everyone fears that the opponent will
- 1:24:34pull ahead.
- 1:24:36>> Yes, I know that I'll be in trouble
- 1:24:38after this podcast, but I will say what
- 1:24:41I think anyway.
- 1:24:42The best solution for everyone is a draw
- 1:24:44.
- 1:24:45Okay? Because any unrestricted power is
- 1:24:48too dangerous for the world.
- 1:24:51In 59 years, I haven't seen a single
- 1:24:54instance of China invading anywhere or
- 1:24:57abusing force. And there is a lot of
- 1:24:59evidence that the Chinese government
- 1:25:01wants prosperity for the global economy
- 1:25:04so that we buy more from them and their
- 1:25:06people become wealthier. Show me
- 1:25:08evidence to the contrary, and I will
- 1:25:10start to be afraid.
- 1:25:12So no, I don't think America should win
- 1:25:14.
- 1:25:15I also don't think China should win, and
- 1:25:17I don't think we need to compete at all
- 1:25:20. The current world order is a scarcity
- 1:25:24mindset, which says:" If they win, I
- 1:25:27lose. If I win, they lose. "That is no
- 1:25:31longer true. With the advent of an
- 1:25:33abundance of intelligence available to
- 1:25:35everyone, it is possible to build a
- 1:25:38world where no one will ever need
- 1:25:40anything. Ever. After all, with
- 1:25:43nanophysics, we are capable of creating
- 1:25:45things literally out of thin air.
- 1:25:48Yes, we can cure any disease, we can
- 1:25:50even effectively harness solar energy
- 1:25:53anywhere in the world.
- 1:25:56We have the power to eliminate the need
- 1:25:58for transportation, because it is
- 1:26:00realistic to create virtual reality so
- 1:26:02convincing that I could sit at home and
- 1:26:04have this conversation with you, and the
- 1:26:06sensations would be exactly the same.
- 1:26:09So much is available to us. And yet we
- 1:26:11build drones, you know?
- 1:26:14And we do this because of a scarcity
- 1:26:16mindset, which says that if they have an
- 1:26:18advantage, I must not fall behind.
- 1:26:21In human history, there have been
- 1:26:23examples like the space station or the
- 1:26:26particle accelerator at CERN, where
- 1:26:28people believed so much in what I call
- 1:26:30mutually assured prosperity.
- 1:26:33Yes, instead of the mutually assured
- 1:26:36destruction that threatens us with
- 1:26:39nuclear confrontation, there is mutually
- 1:26:42assured prosperity.
- 1:26:44If we work together, every person on the
- 1:26:46planet will have everything they need
- 1:26:48endlessly and forever.
- 1:26:50Without any pollution. We are already
- 1:26:52there. I'm ready to argue with anyone. I
- 1:26:54believe that today I can borrow 100 IQ
- 1:26:57points from AI if I truly put my mind to
- 1:27:00it. And learn at an intelligence level
- 1:27:02100 points higher than one's own. You
- 1:27:04know? In four, in three, or maybe in 2
- 1:27:07years, I'll be able to hit 400 IQ points
- 1:27:10, and I'll bet on that with anyone.
- 1:27:13When I can get those 400 points, I'll
- 1:27:16start growing gardens where you walk up
- 1:27:18to one tree and pick an apple, walk up
- 1:27:20to another and pluck an iPhone.
- 1:27:24In terms of molecular cost, they aren't
- 1:27:26that different, you see?
- 1:27:28And we are almost at the goal.
- 1:27:32You know, I worked with people like
- 1:27:34Demis Hassabis.
- 1:27:35Well, I mean, we didn't work together,
- 1:27:37but we were at Google at the same time.
- 1:27:39He has a wonderful approach to solving a
- 1:27:41problem and handling everything else.
- 1:27:45Why are we conquering intelligence to
- 1:27:47create even more problems?
- 1:27:50It's, you know, artificial intelligence
- 1:27:52is a force of polarity, but humanity is
- 1:27:54using it for evil.
- 1:27:57This is where I slightly disagree with
- 1:28:01Roman, in that I believe people will
- 1:28:05finish us off before we do.
- 1:28:09There is a probability that if we don't
- 1:28:12change course and say:" Each and every
- 1:28:14super-powerful technology we create must
- 1:28:17make life better for everyone.
- 1:28:21And those who oppose this must be
- 1:28:24contained.
- 1:28:25"There was an episode of the" Diary of a
- 1:28:28CEO "podcast where Roman was a guest,
- 1:28:30and someone said that arrests should be
- 1:28:32made. For example, some of the CEOs need
- 1:28:36to be stopped.
- 1:28:37I don't know why people don't understand
- 1:28:40this and how politicians can fail to
- 1:28:42notice it.
- 1:28:44After all, there are two ways to realize
- 1:28:46that something needs to be done because
- 1:28:49trouble is approaching.
- 1:28:52One is to just wait for the thunder to
- 1:28:54strike.
- 1:28:55And the other is to use mathematics,
- 1:28:57including applied mathematics, to
- 1:28:58estimate the probability of such an
- 1:29:00outcome.
- 1:29:01And I agree with Roman on this. I think
- 1:29:04the chance of an adverse outcome is 99%
- 1:29:06.
- 1:29:08Right?
- 1:29:10The only difference is I believe
- 1:29:12something bad will happen because one
- 1:29:15person might use AI to harm everyone
- 1:29:17else.
- 1:29:19>> I thank you for this amazing
- 1:29:21conversation. I learned a lot. I am sure
- 1:29:26our viewers did too. Would you like to
- 1:29:28say something to our audience in
- 1:29:29conclusion?
- 1:29:30First of all, you were very kind to
- 1:29:32invite me. It turned out to be a
- 1:29:34pleasant conversation. I hope I haven't
- 1:29:36alarmed the viewers. I am sincerely
- 1:29:38convinced that we all have a choice. You
- 1:29:41know? Everyone has a choice. And it
- 1:29:44starts with awareness.
- 1:29:46It starts with knowing the real truth,
- 1:29:48not what you are told.
- 1:29:51And if you take that step and become
- 1:29:54informed, I must say that I sincerely
- 1:29:57believe humanity is just one decision
- 1:30:01away from an absolute utopia.
- 1:30:05My point of view is that without action
- 1:30:08, we are in for a few years of suffering
- 1:30:11, 10-12, who knows. But if we start
- 1:30:14getting involved, in the end, disaster
- 1:30:16can be avoided. At the same time, the
- 1:30:19end result will be a utopia.
- 1:30:21A world of abundance will come, where
- 1:30:23everything is easy for everyone who
- 1:30:25lives in it. I just don’t see a reason
- 1:30:28why we need to suffer in the process.
- 1:30:30And I think the answer is that the
- 1:30:32decision is up to each of us.
- 1:30:35So start acting accordingly, instead of
- 1:30:38just hoping for change.
About this transcript
This page contains the full transcript of ИИ ЗАБЕРЁТ твою работу к 2027 | Экс-директор Google X by Глеб Соломин, generated from the public captions YouTube serves with the video. The transcript has 13,381 words across 2,225 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.
What you can do with it
Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to choose a caption track available on the video.
Free YouTube transcript tool
YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, caption language selection, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.