Transcript

00:00 The jogan Experience proven right that they they have invented moves. Ai's invented moves that have now been implemented by humans right in a in a very complex game that they never thought that AI was going to be able to be because it requires so much creativity right uh. Arthur we're not quite there but we will be there and by 2029 uh it will match any person that's it 2029. That's just a few years away this yeah well. I'm actually considered conservative people think that will happen like next year the year after. But uh I actually said that in 1999 I said we would uh match any person. By 2029. So 30 years people thought that was totally crazy uh and in fact Stanford had a uh a a conference. They invited several hundred people from around the world world to talk about my prediction and people came in and they and people thought that this would happen but not. By 20129. They thought it would take a hundred years. Yeah I've heard that I've heard that but I think people are amending those uh is it because human beings have a very difficult time grasping the concept of exponential growth. That's exactly right um in fact still economists have a linear View and if you say well it's going to grow exponentially say yeah but maybe 2% a year um. It actually doubles in 14 years.

01:40 Uh and I I brought a chart I can show you that okay that really illustrates. This is this chart available online so we could show people yeah. It's in the book but is it available online that chart where jimy could pull it up and someone could see it just so the folks watching the podcast could see it too. But I could just hold it up to the camera pull it up on pictures. They sent what's it called. What's the title of it. Uh it says uh price performance of computation 1939 to 2023 I have that you have it okay. Great Jam has it yeah. The this The Climb is insane. It's like uh San. What's interesting is that that's an exponential curve and a straight line represents exponential growth and that's absolute straight line for 80 years uh the very first point uh. This is the speed of computers. It was 0.0 7 calculations per second per constant dollar. The last point is 35 billion calculations per second. So there a 20 quadrillion fold increase in those 80 years. But the the speed with which it it gained is is actually the same throughout the entire 80 years because if it was sometimes better and sometimes worse this curve would would bend it would bend up and down.

03:18 It's really very much a straight line uh so the speed with which we increased. It was the same regardless of the technology used and the technology was radically different at the beginning versus the end and yet it it. It increased the speed exactly the same for 80 years. In fact the first 40 years nobody even knew this was happening so it's not like somebody was in charge and saying okay. Next year we have to get to here and people would try to match that we didn't even know. This was happening for 40 years. 40 years later I noticed this for VAR reasons I predicted. It would stay the same the same speed increase each year which which which it has in fact we just put the last do like two weeks ago and it's exactly where it should be so te technology and computation is certainly Prime form of Technology. Uh increases at the same speed and this goes through War and Peace you might say well maybe it's greater during war. No. It's exactly the same you can't tell when there's war or peace or or anything else.On here. It just matches uh from one type of technology to the next uh and it's also true of other things like uh for example getting energy from the Sun uh that's also exponential. It's also just like this uh. It's increased um. We we now are getting uh about a thousand times as much uh energy from the Sun that we did 20 years ago because the implementation of solar panels and the like yeah has the the function of it increased exponentially as well. The function of because what I had understood was that there was a bottleneck in the technology as far as how much you could extract from the Sun from those panels. No not at all. No I mean it's. It's increased uh 99.7% since we started right uh and it's it does the same every year. It's an exponential curve and if if you look at the curve we'll be getting 100% of all the energy we need in 10 years.

05:47 The person who told me that was Elon and Elon was telling me that this is the reason why you can't have a fully solar powerered electric car because it's not capable of absorbing that much from the Sun with a small panel. Like that. He said there's a physical limitation in the panel size. No I mean it's increased 99.7% since we started since what year uh this about um 35 years ago 35 years ago and 99% 99% of the ability of it as well as the expansion of use um I mean you might have to store it. We're also making exponential gains in storage of electricity right technology um. So you don't have to get it all from a solar panel that fits in a car well.The the concept was like could you make a solar paneled car a car that has solar panels on the roof and would that be enough to power the car and he said no he said. It's just not really there yet right. It's not there yet but it it will be there in 10 years. You think so yeah. He he seemed to doubt that he thought that there's a certain limitation of the amount of energy you can get from the Sun period how much it gives out and how much those solar panels can absorb well. You're not going to be able to get it all from the solar panel that fits in a car you're going to have to store some of that energy. Right would the so you wouldn't just be able to drive indefinitely on solar power yeah. That was what he was saying so but you can obviously power a house and especially if you have a roof like. Tesla has those solar pounded roofs now but you can also store the energy for a car um.

07:30 I mean we're we're going to go go to all renewable energy wind and and and Sun uh within 10 years including our ability to store the energy all renewable in 10 years. So what are they going to do with all these nuclear plants and coal power plants and all these things that's completely unnecessary people say we need uh nuclear power which we don't. I you can get it all from the Sun and and wind uh within 10 years you so in 10 years. You' be able to power Los Angeles with sun and wind yes. Really yeah. I I was not aware that we were anywhere near that kind of timeline well that's because people are not taking into account exponential growth so the exponential growth also of the grid because just to pull the amount of power that you would need to charge you know x amount of million if everyone has an electric vehicle by 2035. Let's say then the just the amount of change you would need on the grid would be pretty substantial well. We're making exponential gains on that as well are we yeah yeah I wasn't aware um I i' I had this uh impression that there was a problem with that and especially in Los Angeles. They they've actually asked people at certain times when it's hot out to not charge your car.

08:52 They're not looking at the future that's true now. But it's growing exponentially in every in every field of Technology. Then essentially yeah. Um is the bottleneck a Battery Technology and how how close are they to solving some of these problems with like conflict minerals and the things that we need in order to power these batteries uh I mean our ability to store energy is also growing exponentially so putting all that together uh we'll be able to power. Everything we need within 10 years. Wow most people don't think that so you're you're you're thinking that based on this idea that people have a imagine that computation would grow like this. It's just continuing to do that um and so we have large language models. For example. No one expected that to happen like five years ago right and we had them two years ago. But they didn't work very well. So it began a little less than two years ago that we could actually do large language models uh and and that was very much a surprise to everybody uh so that that's probably the primary example of exponential growth.

10:08 We had Sam Alman on one of the things that he and I were talking about was that AI figured out a way to lie that they used AI to go through a capture system and the AI told the system that it was vision impaired which is not technically a lie. But it used it to bypass are you a robot well. We don't know now is for large language models to say they don't know something so you ask it a question and if that the answer to that question is not in the system it still comes up with an answer. So it'll look at everything and give you its best answer and if the the best answer is not there it's still gives you an answer. But that's uh considered a hallucination and we know a hallucination yeah. That's what it's so AI hallucination so they cannot be wrong. They have to be a to we're actually working on being able to tell if it doesn't know something so if you ask it something and say oh I I don't know that right. Now it can't do that oh wow that's interesting so it. It gives you some answer um and if the answer's not there it just like makes something up. It's the best answer but the best answer isn't very good because it doesn't know the answer and the way to fix hallucinations is to actually give it more more capabilities to memorize things and and give it more information so it knows the answer to it. If you. If you tell uh an answer to a question. It will remember that and give you that correct answer um but these models are not.

11:50 We don't know everything and it has to we have to be able to scan an answer to every single question. Uh which we can't quite do it' be actually better if it could actually answer well. She I don't know that right like in particular like say when it comes to um exploration of the universe. If there's a certain amount of I mean vast amount of the universe we have not explored so if it has to answer questions about that it would just come up with an answer right it'll just come up with an answer which will likely be wrong that's interesting but that that would be a real problem.If someone was counting on the AI to have a solution for something too soon right right. They they don't know everything uh search engines actually know are pretty well vetted and if it actually answers something it'll it's usually correct um unless it's curated. But large language models don't have that capability uh so it' be good actually if they knew that they were wrong that also tell us what we have to fix what about the the idea that AI models are influenced by ideology that AI models have been programmed with certain ideologies. I mean they do learn from people yeah and people have ideologies some of which are some of which are not correct and and that's a large way in which uh it will make things up because it's learning from people um right so right now. Uh if somebody has access to a good uh search engine uh they will check before they actually answer something with a search engine to make sure that it's correct because search engines are generally uh much more accurate generally right when it comes to this idea that people enter information into a computer and then the computer relies on that ideology. Do you anticipate that with artificial general intelligence it'll be agnostic to ideology that it'll be able to reach a point where instead of deciding things based on social norms or whatever the culture is accepted currently that it would look at things more objectively and rationally well eventually eventually. But we still call it artificial general intelligence even if it didn't do that and people certainly do are influenced uh by uh whatever their people that they respect uh feel is correct um and will be as influenced by as people are um and we'll still call it artificial general intelligence. MH um we are starting to check uh what large language models come up with with search engines and that's actually making they're more correct.