1 00:00:00,160 --> 00:00:03,270 A single studio Jibli style film can 2 00:00:03,270 --> 00:00:03,280 A single studio Jibli style film can 3 00:00:03,280 --> 00:00:06,390 A single studio Jibli style film can take 5 to seven years to complete and 4 00:00:06,390 --> 00:00:06,400 take 5 to seven years to complete and 5 00:00:06,400 --> 00:00:10,629 take 5 to seven years to complete and require over a 100,000 handdrawn frames. 6 00:00:10,629 --> 00:00:10,639 require over a 100,000 handdrawn frames. 7 00:00:10,639 --> 00:00:13,110 require over a 100,000 handdrawn frames. But AI can generate a frame like this in 8 00:00:13,110 --> 00:00:13,120 But AI can generate a frame like this in 9 00:00:13,120 --> 00:00:16,150 But AI can generate a frame like this in under 10 seconds. In just 5 years, we've 10 00:00:16,150 --> 00:00:16,160 under 10 seconds. In just 5 years, we've 11 00:00:16,160 --> 00:00:19,349 under 10 seconds. In just 5 years, we've gone from clunky chat bots to AI models 12 00:00:19,349 --> 00:00:19,359 gone from clunky chat bots to AI models 13 00:00:19,359 --> 00:00:22,550 gone from clunky chat bots to AI models scoring in the top 10% on the bar exam 14 00:00:22,550 --> 00:00:22,560 scoring in the top 10% on the bar exam 15 00:00:22,560 --> 00:00:25,070 scoring in the top 10% on the bar exam and even helping scientists discover new 16 00:00:25,070 --> 00:00:25,080 and even helping scientists discover new 17 00:00:25,080 --> 00:00:27,589 and even helping scientists discover new antibiotics. We're no longer advancing 18 00:00:27,589 --> 00:00:27,599 antibiotics. We're no longer advancing 19 00:00:27,599 --> 00:00:30,470 antibiotics. We're no longer advancing alongside AI. we're being left behind. 20 00:00:30,470 --> 00:00:30,480 alongside AI. we're being left behind. 21 00:00:30,480 --> 00:00:32,870 alongside AI. we're being left behind. My name is Sam Woods. I've used AI since 22 00:00:32,870 --> 00:00:32,880 My name is Sam Woods. I've used AI since 23 00:00:32,880 --> 00:00:35,830 My name is Sam Woods. I've used AI since 2016, and I'm now training Fortune 1000 24 00:00:35,830 --> 00:00:35,840 2016, and I'm now training Fortune 1000 25 00:00:35,840 --> 00:00:38,470 2016, and I'm now training Fortune 1000 teams on how to use it so that they're 26 00:00:38,470 --> 00:00:38,480 teams on how to use it so that they're 27 00:00:38,480 --> 00:00:40,790 teams on how to use it so that they're prepared for what comes next. I've 28 00:00:40,790 --> 00:00:40,800 prepared for what comes next. I've 29 00:00:40,800 --> 00:00:42,549 prepared for what comes next. I've consulted in over 37 different 30 00:00:42,549 --> 00:00:42,559 consulted in over 37 different 31 00:00:42,559 --> 00:00:45,510 consulted in over 37 different industries, and I'm now convinced of one 32 00:00:45,510 --> 00:00:45,520 industries, and I'm now convinced of one 33 00:00:45,520 --> 00:00:48,310 industries, and I'm now convinced of one thing. Humans may no longer be the 34 00:00:48,310 --> 00:00:48,320 thing. Humans may no longer be the 35 00:00:48,320 --> 00:00:51,270 thing. Humans may no longer be the smartest species on Earth. Right now, AI 36 00:00:51,270 --> 00:00:51,280 smartest species on Earth. Right now, AI 37 00:00:51,280 --> 00:00:53,510 smartest species on Earth. Right now, AI doesn't seem smarter than you, right? It 38 00:00:53,510 --> 00:00:53,520 doesn't seem smarter than you, right? It 39 00:00:53,520 --> 00:00:56,150 doesn't seem smarter than you, right? It doesn't have your intuition. Now AI 40 00:00:56,150 --> 00:00:56,160 doesn't have your intuition. Now AI 41 00:00:56,160 --> 00:00:59,830 doesn't have your intuition. Now AI lacks common sense, intuition, instinct. 42 00:00:59,830 --> 00:00:59,840 lacks common sense, intuition, instinct. 43 00:00:59,840 --> 00:01:01,670 lacks common sense, intuition, instinct. It cannot understand why something 44 00:01:01,670 --> 00:01:01,680 It cannot understand why something 45 00:01:01,680 --> 00:01:04,469 It cannot understand why something happens and it can't really adapt when 46 00:01:04,469 --> 00:01:04,479 happens and it can't really adapt when 47 00:01:04,479 --> 00:01:07,030 happens and it can't really adapt when rules change. And that's because AI 48 00:01:07,030 --> 00:01:07,040 rules change. And that's because AI 49 00:01:07,040 --> 00:01:09,830 rules change. And that's because AI recombines existing data which means 50 00:01:09,830 --> 00:01:09,840 recombines existing data which means 51 00:01:09,840 --> 00:01:12,710 recombines existing data which means that it cannot imagine, think or be 52 00:01:12,710 --> 00:01:12,720 that it cannot imagine, think or be 53 00:01:12,720 --> 00:01:15,030 that it cannot imagine, think or be creative in the way humans are. But that 54 00:01:15,030 --> 00:01:15,040 creative in the way humans are. But that 55 00:01:15,040 --> 00:01:17,350 creative in the way humans are. But that creates a false sense of comfort because 56 00:01:17,350 --> 00:01:17,360 creates a false sense of comfort because 57 00:01:17,360 --> 00:01:19,990 creates a false sense of comfort because we assume intelligence must look human 58 00:01:19,990 --> 00:01:20,000 we assume intelligence must look human 59 00:01:20,000 --> 00:01:23,190 we assume intelligence must look human to be powerful. And that's a dangerous 60 00:01:23,190 --> 00:01:23,200 to be powerful. And that's a dangerous 61 00:01:23,200 --> 00:01:25,830 to be powerful. And that's a dangerous illusion. As Eleazar Yudkowski, the 62 00:01:25,830 --> 00:01:25,840 illusion. As Eleazar Yudkowski, the 63 00:01:25,840 --> 00:01:28,550 illusion. As Eleazar Yudkowski, the co-founder and research fellow at the 64 00:01:28,550 --> 00:01:28,560 co-founder and research fellow at the 65 00:01:28,560 --> 00:01:30,870 co-founder and research fellow at the Machine Intelligence Research Institute 66 00:01:30,870 --> 00:01:30,880 Machine Intelligence Research Institute 67 00:01:30,880 --> 00:01:33,830 Machine Intelligence Research Institute said, and I quote, "By far the greatest 68 00:01:33,830 --> 00:01:33,840 said, and I quote, "By far the greatest 69 00:01:33,840 --> 00:01:36,069 said, and I quote, "By far the greatest danger of artificial intelligence is 70 00:01:36,069 --> 00:01:36,079 danger of artificial intelligence is 71 00:01:36,079 --> 00:01:39,350 danger of artificial intelligence is that people conclude too early that they 72 00:01:39,350 --> 00:01:39,360 that people conclude too early that they 73 00:01:39,360 --> 00:01:42,230 that people conclude too early that they understand it." So, humanity took 74 00:01:42,230 --> 00:01:42,240 understand it." So, humanity took 75 00:01:42,240 --> 00:01:44,550 understand it." So, humanity took centuries to master the complex systems 76 00:01:44,550 --> 00:01:44,560 centuries to master the complex systems 77 00:01:44,560 --> 00:01:46,710 centuries to master the complex systems that we live by today. Mathematics 78 00:01:46,710 --> 00:01:46,720 that we live by today. Mathematics 79 00:01:46,720 --> 00:01:49,190 that we live by today. Mathematics evolved over thousands of years from 80 00:01:49,190 --> 00:01:49,200 evolved over thousands of years from 81 00:01:49,200 --> 00:01:51,190 evolved over thousands of years from Babylonian arithmetic all the way to 82 00:01:51,190 --> 00:01:51,200 Babylonian arithmetic all the way to 83 00:01:51,200 --> 00:01:54,230 Babylonian arithmetic all the way to Newton's calculus in the 1600s. Language 84 00:01:54,230 --> 00:01:54,240 Newton's calculus in the 1600s. Language 85 00:01:54,240 --> 00:01:57,510 Newton's calculus in the 1600s. Language itself appeared around 3000 B.CE. But 86 00:01:57,510 --> 00:01:57,520 itself appeared around 3000 B.CE. But 87 00:01:57,520 --> 00:01:59,830 itself appeared around 3000 B.CE. But modern grammar and linguistics theory 88 00:01:59,830 --> 00:01:59,840 modern grammar and linguistics theory 89 00:01:59,840 --> 00:02:02,109 modern grammar and linguistics theory only began forming in the 19th and 20th 90 00:02:02,109 --> 00:02:02,119 only began forming in the 19th and 20th 91 00:02:02,119 --> 00:02:05,030 only began forming in the 19th and 20th centuries. And art, everything from 92 00:02:05,030 --> 00:02:05,040 centuries. And art, everything from 93 00:02:05,040 --> 00:02:06,789 centuries. And art, everything from perspective drawing to classical 94 00:02:06,789 --> 00:02:06,799 perspective drawing to classical 95 00:02:06,799 --> 00:02:10,070 perspective drawing to classical composition took millennia of cultural 96 00:02:10,070 --> 00:02:10,080 composition took millennia of cultural 97 00:02:10,080 --> 00:02:13,750 composition took millennia of cultural evolution. But today, AI can replicate 98 00:02:13,750 --> 00:02:13,760 evolution. But today, AI can replicate 99 00:02:13,760 --> 00:02:16,710 evolution. But today, AI can replicate the handdrawn magic of Studio Ghibli in 100 00:02:16,710 --> 00:02:16,720 the handdrawn magic of Studio Ghibli in 101 00:02:16,720 --> 00:02:19,350 the handdrawn magic of Studio Ghibli in a style that took decades of human 102 00:02:19,350 --> 00:02:19,360 a style that took decades of human 103 00:02:19,360 --> 00:02:22,790 a style that took decades of human mastery to develop in just seconds. 104 00:02:22,790 --> 00:02:22,800 mastery to develop in just seconds. 105 00:02:22,800 --> 00:02:25,910 mastery to develop in just seconds. Let's also take Deep Mice Alph Go. Go is 106 00:02:25,910 --> 00:02:25,920 Let's also take Deep Mice Alph Go. Go is 107 00:02:25,920 --> 00:02:28,390 Let's also take Deep Mice Alph Go. Go is a game so complex it's been described as 108 00:02:28,390 --> 00:02:28,400 a game so complex it's been described as 109 00:02:28,400 --> 00:02:31,190 a game so complex it's been described as the pinnacle of human strategy. It took 110 00:02:31,190 --> 00:02:31,200 the pinnacle of human strategy. It took 111 00:02:31,200 --> 00:02:34,710 the pinnacle of human strategy. It took us over 2500 years to refine it. Alph Go 112 00:02:34,710 --> 00:02:34,720 us over 2500 years to refine it. Alph Go 113 00:02:34,720 --> 00:02:37,910 us over 2500 years to refine it. Alph Go learned it in 3 days. Then it went on to 114 00:02:37,910 --> 00:02:37,920 learned it in 3 days. Then it went on to 115 00:02:37,920 --> 00:02:40,150 learned it in 3 days. Then it went on to defeat the world's best human player, 116 00:02:40,150 --> 00:02:40,160 defeat the world's best human player, 117 00:02:40,160 --> 00:02:42,509 defeat the world's best human player, not just with brute force, but with 118 00:02:42,509 --> 00:02:42,519 not just with brute force, but with 119 00:02:42,519 --> 00:02:45,670 not just with brute force, but with creativity. It made moves no human had 120 00:02:45,670 --> 00:02:45,680 creativity. It made moves no human had 121 00:02:45,680 --> 00:02:49,309 creativity. It made moves no human had imagined before. Or look at OpenAI's 122 00:02:49,309 --> 00:02:49,319 imagined before. Or look at OpenAI's 123 00:02:49,319 --> 00:02:53,430 imagined before. Or look at OpenAI's GBT40, 01, 03, 04, and all these 124 00:02:53,430 --> 00:02:53,440 GBT40, 01, 03, 04, and all these 125 00:02:53,440 --> 00:02:56,710 GBT40, 01, 03, 04, and all these reasoning models. They can write essays, 126 00:02:56,710 --> 00:02:56,720 reasoning models. They can write essays, 127 00:02:56,720 --> 00:02:59,270 reasoning models. They can write essays, scripts, software, perform intensive 128 00:02:59,270 --> 00:02:59,280 scripts, software, perform intensive 129 00:02:59,280 --> 00:03:01,589 scripts, software, perform intensive data analysis, and generate business 130 00:03:01,589 --> 00:03:01,599 data analysis, and generate business 131 00:03:01,599 --> 00:03:03,910 data analysis, and generate business strategies out of thin air. What takes a 132 00:03:03,910 --> 00:03:03,920 strategies out of thin air. What takes a 133 00:03:03,920 --> 00:03:06,509 strategies out of thin air. What takes a human a specialized degree and years of 134 00:03:06,509 --> 00:03:06,519 human a specialized degree and years of 135 00:03:06,519 --> 00:03:09,430 human a specialized degree and years of experience, these AI models can generate 136 00:03:09,430 --> 00:03:09,440 experience, these AI models can generate 137 00:03:09,440 --> 00:03:11,910 experience, these AI models can generate instantly without ever getting tired, 138 00:03:11,910 --> 00:03:11,920 instantly without ever getting tired, 139 00:03:11,920 --> 00:03:14,869 instantly without ever getting tired, distracted, or emotional. And this isn't 140 00:03:14,869 --> 00:03:14,879 distracted, or emotional. And this isn't 141 00:03:14,879 --> 00:03:16,949 distracted, or emotional. And this isn't a linear progression. This is 142 00:03:16,949 --> 00:03:16,959 a linear progression. This is 143 00:03:16,959 --> 00:03:19,509 a linear progression. This is exponential. The AI you're using today 144 00:03:19,509 --> 00:03:19,519 exponential. The AI you're using today 145 00:03:19,519 --> 00:03:21,830 exponential. The AI you're using today is 100 times more powerful than what we 146 00:03:21,830 --> 00:03:21,840 is 100 times more powerful than what we 147 00:03:21,840 --> 00:03:24,869 is 100 times more powerful than what we had 5 years ago. Think about that. In 148 00:03:24,869 --> 00:03:24,879 had 5 years ago. Think about that. In 149 00:03:24,879 --> 00:03:27,190 had 5 years ago. Think about that. In human terms, it's like going from the 150 00:03:27,190 --> 00:03:27,200 human terms, it's like going from the 151 00:03:27,200 --> 00:03:30,350 human terms, it's like going from the Wright brothers airplane to interstellar 152 00:03:30,350 --> 00:03:30,360 Wright brothers airplane to interstellar 153 00:03:30,360 --> 00:03:34,030 Wright brothers airplane to interstellar rockets in into space in less than a 154 00:03:34,030 --> 00:03:34,040 rockets in into space in less than a 155 00:03:34,040 --> 00:03:36,949 rockets in into space in less than a decade. That's insane. So when people 156 00:03:36,949 --> 00:03:36,959 decade. That's insane. So when people 157 00:03:36,959 --> 00:03:40,149 decade. That's insane. So when people ask me, is AI smarter than humans? 158 00:03:40,149 --> 00:03:40,159 ask me, is AI smarter than humans? 159 00:03:40,159 --> 00:03:41,750 ask me, is AI smarter than humans? They're asking the wrong question. 160 00:03:41,750 --> 00:03:41,760 They're asking the wrong question. 161 00:03:41,760 --> 00:03:43,910 They're asking the wrong question. Intelligence isn't about mimicking us. 162 00:03:43,910 --> 00:03:43,920 Intelligence isn't about mimicking us. 163 00:03:43,920 --> 00:03:45,990 Intelligence isn't about mimicking us. It is about solving problems faster, 164 00:03:45,990 --> 00:03:46,000 It is about solving problems faster, 165 00:03:46,000 --> 00:03:48,149 It is about solving problems faster, cheaper, and at a scale that we can't 166 00:03:48,149 --> 00:03:48,159 cheaper, and at a scale that we can't 167 00:03:48,159 --> 00:03:50,470 cheaper, and at a scale that we can't comprehend. That's why AI is going to 168 00:03:50,470 --> 00:03:50,480 comprehend. That's why AI is going to 169 00:03:50,480 --> 00:03:52,550 comprehend. That's why AI is going to surpass human intelligence in ways we 170 00:03:52,550 --> 00:03:52,560 surpass human intelligence in ways we 171 00:03:52,560 --> 00:03:55,270 surpass human intelligence in ways we aren't ready for. Now when we talk about 172 00:03:55,270 --> 00:03:55,280 aren't ready for. Now when we talk about 173 00:03:55,280 --> 00:03:58,390 aren't ready for. Now when we talk about AI taking over, we imagine an AI that 174 00:03:58,390 --> 00:03:58,400 AI taking over, we imagine an AI that 175 00:03:58,400 --> 00:04:02,309 AI taking over, we imagine an AI that has consciousness. But a study by 19 176 00:04:02,309 --> 00:04:02,319 has consciousness. But a study by 19 177 00:04:02,319 --> 00:04:05,750 has consciousness. But a study by 19 researchers assessed AI systems against 178 00:04:05,750 --> 00:04:05,760 researchers assessed AI systems against 179 00:04:05,760 --> 00:04:07,710 researchers assessed AI systems against neuroscientific theories of 180 00:04:07,710 --> 00:04:07,720 neuroscientific theories of 181 00:04:07,720 --> 00:04:10,309 neuroscientific theories of consciousness such as the global work 182 00:04:10,309 --> 00:04:10,319 consciousness such as the global work 183 00:04:10,319 --> 00:04:13,589 consciousness such as the global work space theory and higher order theories. 184 00:04:13,589 --> 00:04:13,599 space theory and higher order theories. 185 00:04:13,599 --> 00:04:15,910 space theory and higher order theories. In that study found that current AI 186 00:04:15,910 --> 00:04:15,920 In that study found that current AI 187 00:04:15,920 --> 00:04:18,469 In that study found that current AI systems do not exhibit the necessary 188 00:04:18,469 --> 00:04:18,479 systems do not exhibit the necessary 189 00:04:18,479 --> 00:04:21,590 systems do not exhibit the necessary indicators of consciousness. And the 190 00:04:21,590 --> 00:04:21,600 indicators of consciousness. And the 191 00:04:21,600 --> 00:04:23,749 indicators of consciousness. And the neurogenetic structuralism theory 192 00:04:23,749 --> 00:04:23,759 neurogenetic structuralism theory 193 00:04:23,759 --> 00:04:26,790 neurogenetic structuralism theory suggests that consciousness arises from 194 00:04:26,790 --> 00:04:26,800 suggests that consciousness arises from 195 00:04:26,800 --> 00:04:29,790 suggests that consciousness arises from complex physiology of biological 196 00:04:29,790 --> 00:04:29,800 complex physiology of biological 197 00:04:29,800 --> 00:04:33,350 complex physiology of biological neurons. Now AI lacks such a biological 198 00:04:33,350 --> 00:04:33,360 neurons. Now AI lacks such a biological 199 00:04:33,360 --> 00:04:35,749 neurons. Now AI lacks such a biological structure. So it cannot achieve 200 00:04:35,749 --> 00:04:35,759 structure. So it cannot achieve 201 00:04:35,759 --> 00:04:38,150 structure. So it cannot achieve consciousness the way us humans 202 00:04:38,150 --> 00:04:38,160 consciousness the way us humans 203 00:04:38,160 --> 00:04:40,590 consciousness the way us humans understand or think we understand 204 00:04:40,590 --> 00:04:40,600 understand or think we understand 205 00:04:40,600 --> 00:04:42,710 understand or think we understand consciousness. AI may not have 206 00:04:42,710 --> 00:04:42,720 consciousness. AI may not have 207 00:04:42,720 --> 00:04:44,629 consciousness. AI may not have consciousness, but it doesn't need to 208 00:04:44,629 --> 00:04:44,639 consciousness, but it doesn't need to 209 00:04:44,639 --> 00:04:47,909 consciousness, but it doesn't need to think like us to outsmart us. In fact, 210 00:04:47,909 --> 00:04:47,919 think like us to outsmart us. In fact, 211 00:04:47,919 --> 00:04:49,830 think like us to outsmart us. In fact, one of the biggest myths about AI is 212 00:04:49,830 --> 00:04:49,840 one of the biggest myths about AI is 213 00:04:49,840 --> 00:04:52,350 one of the biggest myths about AI is that it needs to be conscious to be 214 00:04:52,350 --> 00:04:52,360 that it needs to be conscious to be 215 00:04:52,360 --> 00:04:54,550 that it needs to be conscious to be dangerous. We like to imagine that 216 00:04:54,550 --> 00:04:54,560 dangerous. We like to imagine that 217 00:04:54,560 --> 00:04:57,590 dangerous. We like to imagine that unless AI can feel or reflect, it can't 218 00:04:57,590 --> 00:04:57,600 unless AI can feel or reflect, it can't 219 00:04:57,600 --> 00:05:00,950 unless AI can feel or reflect, it can't truly think. But that's human arrogance 220 00:05:00,950 --> 00:05:00,960 truly think. But that's human arrogance 221 00:05:00,960 --> 00:05:03,270 truly think. But that's human arrogance disguised as philosophy. Consciousness 222 00:05:03,270 --> 00:05:03,280 disguised as philosophy. Consciousness 223 00:05:03,280 --> 00:05:05,189 disguised as philosophy. Consciousness might be a beautiful feature of the 224 00:05:05,189 --> 00:05:05,199 might be a beautiful feature of the 225 00:05:05,199 --> 00:05:08,550 might be a beautiful feature of the human experience. But it is not a 226 00:05:08,550 --> 00:05:08,560 human experience. But it is not a 227 00:05:08,560 --> 00:05:12,150 human experience. But it is not a prerequisite for dominance. A chess 228 00:05:12,150 --> 00:05:12,160 prerequisite for dominance. A chess 229 00:05:12,160 --> 00:05:14,310 prerequisite for dominance. A chess computer doesn't know that it's playing 230 00:05:14,310 --> 00:05:14,320 computer doesn't know that it's playing 231 00:05:14,320 --> 00:05:17,230 computer doesn't know that it's playing chess. It doesn't daydream about being a 232 00:05:17,230 --> 00:05:17,240 chess. It doesn't daydream about being a 233 00:05:17,240 --> 00:05:19,830 chess. It doesn't daydream about being a grandmaster, but it still wins. And 234 00:05:19,830 --> 00:05:19,840 grandmaster, but it still wins. And 235 00:05:19,840 --> 00:05:21,990 grandmaster, but it still wins. And that's just the tip of the iceberg. In 236 00:05:21,990 --> 00:05:22,000 that's just the tip of the iceberg. In 237 00:05:22,000 --> 00:05:23,870 that's just the tip of the iceberg. In the real world, we're watching 238 00:05:23,870 --> 00:05:23,880 the real world, we're watching 239 00:05:23,880 --> 00:05:27,670 the real world, we're watching non-concious systems quietly outperform 240 00:05:27,670 --> 00:05:27,680 non-concious systems quietly outperform 241 00:05:27,680 --> 00:05:30,390 non-concious systems quietly outperform humans at highstakes tasks. I'm talking 242 00:05:30,390 --> 00:05:30,400 humans at highstakes tasks. I'm talking 243 00:05:30,400 --> 00:05:33,270 humans at highstakes tasks. I'm talking about medical diagnosis, legal analysis, 244 00:05:33,270 --> 00:05:33,280 about medical diagnosis, legal analysis, 245 00:05:33,280 --> 00:05:36,230 about medical diagnosis, legal analysis, and even military simulations. This is a 246 00:05:36,230 --> 00:05:36,240 and even military simulations. This is a 247 00:05:36,240 --> 00:05:38,629 and even military simulations. This is a major reason DARPA is testing autonomous 248 00:05:38,629 --> 00:05:38,639 major reason DARPA is testing autonomous 249 00:05:38,639 --> 00:05:41,909 major reason DARPA is testing autonomous AI weapons. I mean, machines obviously 250 00:05:41,909 --> 00:05:41,919 AI weapons. I mean, machines obviously 251 00:05:41,919 --> 00:05:43,909 AI weapons. I mean, machines obviously won't hesitate. They'll just do what 252 00:05:43,909 --> 00:05:43,919 won't hesitate. They'll just do what 253 00:05:43,919 --> 00:05:45,909 won't hesitate. They'll just do what they're told and obey your command. 254 00:05:45,909 --> 00:05:45,919 they're told and obey your command. 255 00:05:45,919 --> 00:05:48,550 they're told and obey your command. Tools like Co- Counsel and Harvey AI are 256 00:05:48,550 --> 00:05:48,560 Tools like Co- Counsel and Harvey AI are 257 00:05:48,560 --> 00:05:50,790 Tools like Co- Counsel and Harvey AI are already being used by major law firms to 258 00:05:50,790 --> 00:05:50,800 already being used by major law firms to 259 00:05:50,800 --> 00:05:52,870 already being used by major law firms to draft legal arguments, summarize 260 00:05:52,870 --> 00:05:52,880 draft legal arguments, summarize 261 00:05:52,880 --> 00:05:55,270 draft legal arguments, summarize depositions, and analyze contracts 262 00:05:55,270 --> 00:05:55,280 depositions, and analyze contracts 263 00:05:55,280 --> 00:05:57,909 depositions, and analyze contracts faster and more accurately than junior 264 00:05:57,909 --> 00:05:57,919 faster and more accurately than junior 265 00:05:57,919 --> 00:06:01,110 faster and more accurately than junior attorneys. In 2020, an AI called 266 00:06:01,110 --> 00:06:01,120 attorneys. In 2020, an AI called 267 00:06:01,120 --> 00:06:03,830 attorneys. In 2020, an AI called Alphafold figured out proteins fold, 268 00:06:03,830 --> 00:06:03,840 Alphafold figured out proteins fold, 269 00:06:03,840 --> 00:06:05,590 Alphafold figured out proteins fold, something that scientists couldn't solve 270 00:06:05,590 --> 00:06:05,600 something that scientists couldn't solve 271 00:06:05,600 --> 00:06:08,070 something that scientists couldn't solve for 50 years, and that helped with 272 00:06:08,070 --> 00:06:08,080 for 50 years, and that helped with 273 00:06:08,080 --> 00:06:10,309 for 50 years, and that helped with making new medicines. Another AI, 274 00:06:10,309 --> 00:06:10,319 making new medicines. Another AI, 275 00:06:10,319 --> 00:06:13,270 making new medicines. Another AI, Libratas, beat top poker players by 276 00:06:13,270 --> 00:06:13,280 Libratas, beat top poker players by 277 00:06:13,280 --> 00:06:15,350 Libratas, beat top poker players by learning how to bluff and change his 278 00:06:15,350 --> 00:06:15,360 learning how to bluff and change his 279 00:06:15,360 --> 00:06:17,749 learning how to bluff and change his strategy. Something even smart people 280 00:06:17,749 --> 00:06:17,759 strategy. Something even smart people 281 00:06:17,759 --> 00:06:21,110 strategy. Something even smart people struggle with. And in 2023, an AI system 282 00:06:21,110 --> 00:06:21,120 struggle with. And in 2023, an AI system 283 00:06:21,120 --> 00:06:23,749 struggle with. And in 2023, an AI system beat top radiologist at identifying 284 00:06:23,749 --> 00:06:23,759 beat top radiologist at identifying 285 00:06:23,759 --> 00:06:25,909 beat top radiologist at identifying breast cancer by spotting patterns 286 00:06:25,909 --> 00:06:25,919 breast cancer by spotting patterns 287 00:06:25,919 --> 00:06:28,230 breast cancer by spotting patterns invisible to the human eye. Not by 288 00:06:28,230 --> 00:06:28,240 invisible to the human eye. Not by 289 00:06:28,240 --> 00:06:30,710 invisible to the human eye. Not by understanding what cancer is, but by 290 00:06:30,710 --> 00:06:30,720 understanding what cancer is, but by 291 00:06:30,720 --> 00:06:33,309 understanding what cancer is, but by pure pattern recognition. That's cold, 292 00:06:33,309 --> 00:06:33,319 pure pattern recognition. That's cold, 293 00:06:33,319 --> 00:06:36,390 pure pattern recognition. That's cold, relentless logic. The kind that doesn't 294 00:06:36,390 --> 00:06:36,400 relentless logic. The kind that doesn't 295 00:06:36,400 --> 00:06:38,550 relentless logic. The kind that doesn't second guessess. AI also doesn't fall 296 00:06:38,550 --> 00:06:38,560 second guessess. AI also doesn't fall 297 00:06:38,560 --> 00:06:40,950 second guessess. AI also doesn't fall prey to ego. It doesn't sleep on 298 00:06:40,950 --> 00:06:40,960 prey to ego. It doesn't sleep on 299 00:06:40,960 --> 00:06:43,510 prey to ego. It doesn't sleep on decisions. And most importantly, it 300 00:06:43,510 --> 00:06:43,520 decisions. And most importantly, it 301 00:06:43,520 --> 00:06:45,510 decisions. And most importantly, it learns from trillions of data points in 302 00:06:45,510 --> 00:06:45,520 learns from trillions of data points in 303 00:06:45,520 --> 00:06:47,590 learns from trillions of data points in a fraction of the time. While we're 304 00:06:47,590 --> 00:06:47,600 a fraction of the time. While we're 305 00:06:47,600 --> 00:06:49,590 a fraction of the time. While we're still debating whether AI can be 306 00:06:49,590 --> 00:06:49,600 still debating whether AI can be 307 00:06:49,600 --> 00:06:52,309 still debating whether AI can be conscious, it's already making decisions 308 00:06:52,309 --> 00:06:52,319 conscious, it's already making decisions 309 00:06:52,319 --> 00:06:54,710 conscious, it's already making decisions faster, cheaper, better, and without 310 00:06:54,710 --> 00:06:54,720 faster, cheaper, better, and without 311 00:06:54,720 --> 00:06:57,189 faster, cheaper, better, and without bias fatigue. So even without actually 312 00:06:57,189 --> 00:06:57,199 bias fatigue. So even without actually 313 00:06:57,199 --> 00:07:01,270 bias fatigue. So even without actually being human, AI could beat humans at 314 00:07:01,270 --> 00:07:01,280 being human, AI could beat humans at 315 00:07:01,280 --> 00:07:04,309 being human, AI could beat humans at making decisions in every domain. Ever 316 00:07:04,309 --> 00:07:04,319 making decisions in every domain. Ever 317 00:07:04,319 --> 00:07:06,110 making decisions in every domain. Ever since AI entered the public 318 00:07:06,110 --> 00:07:06,120 since AI entered the public 319 00:07:06,120 --> 00:07:08,230 since AI entered the public consciousness, the default fear has been 320 00:07:08,230 --> 00:07:08,240 consciousness, the default fear has been 321 00:07:08,240 --> 00:07:10,070 consciousness, the default fear has been the same. What happens when machines 322 00:07:10,070 --> 00:07:10,080 the same. What happens when machines 323 00:07:10,080 --> 00:07:12,390 the same. What happens when machines take our jobs? The people closest to 324 00:07:12,390 --> 00:07:12,400 take our jobs? The people closest to 325 00:07:12,400 --> 00:07:15,029 take our jobs? The people closest to AI's development aren't just worried 326 00:07:15,029 --> 00:07:15,039 AI's development aren't just worried 327 00:07:15,039 --> 00:07:15,790 AI's development aren't just worried about 328 00:07:15,790 --> 00:07:15,800 about 329 00:07:15,800 --> 00:07:19,189 about replacement. The deeper threat is that 330 00:07:19,189 --> 00:07:19,199 replacement. The deeper threat is that 331 00:07:19,199 --> 00:07:21,830 replacement. The deeper threat is that it surpasses us permanently. For 332 00:07:21,830 --> 00:07:21,840 it surpasses us permanently. For 333 00:07:21,840 --> 00:07:24,230 it surpasses us permanently. For example, Elon Musk famously said on X 334 00:07:24,230 --> 00:07:24,240 example, Elon Musk famously said on X 335 00:07:24,240 --> 00:07:26,629 example, Elon Musk famously said on X that AI is potentially more dangerous 336 00:07:26,629 --> 00:07:26,639 that AI is potentially more dangerous 337 00:07:26,639 --> 00:07:29,830 that AI is potentially more dangerous than nukes. Jeffrey Hinton, often known 338 00:07:29,830 --> 00:07:29,840 than nukes. Jeffrey Hinton, often known 339 00:07:29,840 --> 00:07:32,710 than nukes. Jeffrey Hinton, often known as the godfather of AI, said in 2023 340 00:07:32,710 --> 00:07:32,720 as the godfather of AI, said in 2023 341 00:07:32,720 --> 00:07:35,670 as the godfather of AI, said in 2023 that there's a 5050 chance that 342 00:07:35,670 --> 00:07:35,680 that there's a 5050 chance that 343 00:07:35,680 --> 00:07:37,589 that there's a 5050 chance that artificial intelligence will surpass 344 00:07:37,589 --> 00:07:37,599 artificial intelligence will surpass 345 00:07:37,599 --> 00:07:41,110 artificial intelligence will surpass human intelligence within 20 years. Some 346 00:07:41,110 --> 00:07:41,120 human intelligence within 20 years. Some 347 00:07:41,120 --> 00:07:43,670 human intelligence within 20 years. Some say that's going to happen faster. And 348 00:07:43,670 --> 00:07:43,680 say that's going to happen faster. And 349 00:07:43,680 --> 00:07:45,830 say that's going to happen faster. And futurists like Ray Kerszswe has a 350 00:07:45,830 --> 00:07:45,840 futurists like Ray Kerszswe has a 351 00:07:45,840 --> 00:07:47,909 futurists like Ray Kerszswe has a different name for this. He calls it the 352 00:07:47,909 --> 00:07:47,919 different name for this. He calls it the 353 00:07:47,919 --> 00:07:49,430 different name for this. He calls it the singularity. 354 00:07:49,430 --> 00:07:49,440 singularity. 355 00:07:49,440 --> 00:07:51,589 singularity. It's a moment where humans and machines 356 00:07:51,589 --> 00:07:51,599 It's a moment where humans and machines 357 00:07:51,599 --> 00:07:54,469 It's a moment where humans and machines become so deeply integrated that the 358 00:07:54,469 --> 00:07:54,479 become so deeply integrated that the 359 00:07:54,479 --> 00:07:56,710 become so deeply integrated that the line between biology and technology 360 00:07:56,710 --> 00:07:56,720 line between biology and technology 361 00:07:56,720 --> 00:07:58,869 line between biology and technology disappears completely. It sounds 362 00:07:58,869 --> 00:07:58,879 disappears completely. It sounds 363 00:07:58,879 --> 00:08:01,110 disappears completely. It sounds absolutely wild until you realize that 364 00:08:01,110 --> 00:08:01,120 absolutely wild until you realize that 365 00:08:01,120 --> 00:08:03,670 absolutely wild until you realize that it's already in motion. In 2024, 366 00:08:03,670 --> 00:08:03,680 it's already in motion. In 2024, 367 00:08:03,680 --> 00:08:06,070 it's already in motion. In 2024, Neuralink successfully implanted a brain 368 00:08:06,070 --> 00:08:06,080 Neuralink successfully implanted a brain 369 00:08:06,080 --> 00:08:08,950 Neuralink successfully implanted a brain chip in his first human patient, a 370 00:08:08,950 --> 00:08:08,960 chip in his first human patient, a 371 00:08:08,960 --> 00:08:11,430 chip in his first human patient, a 30-year-old quadripollegic man named 372 00:08:11,430 --> 00:08:11,440 30-year-old quadripollegic man named 373 00:08:11,440 --> 00:08:14,550 30-year-old quadripollegic man named Nolan Arbomb, who was suddenly able to 374 00:08:14,550 --> 00:08:14,560 Nolan Arbomb, who was suddenly able to 375 00:08:14,560 --> 00:08:18,469 Nolan Arbomb, who was suddenly able to move the computer cursor with only his 376 00:08:18,469 --> 00:08:18,479 move the computer cursor with only his 377 00:08:18,479 --> 00:08:21,029 move the computer cursor with only his thoughts. So, the future might not be AI 378 00:08:21,029 --> 00:08:21,039 thoughts. So, the future might not be AI 379 00:08:21,039 --> 00:08:24,469 thoughts. So, the future might not be AI versus humans, but AI augmented humans 380 00:08:24,469 --> 00:08:24,479 versus humans, but AI augmented humans 381 00:08:24,479 --> 00:08:27,270 versus humans, but AI augmented humans versus extinction. That sounds like 382 00:08:27,270 --> 00:08:27,280 versus extinction. That sounds like 383 00:08:27,280 --> 00:08:30,390 versus extinction. That sounds like science fiction, but that is a plausible 384 00:08:30,390 --> 00:08:30,400 science fiction, but that is a plausible 385 00:08:30,400 --> 00:08:33,029 science fiction, but that is a plausible outcome on our current trajectory. When 386 00:08:33,029 --> 00:08:33,039 outcome on our current trajectory. When 387 00:08:33,039 --> 00:08:35,110 outcome on our current trajectory. When we stop thinking like it's us versus 388 00:08:35,110 --> 00:08:35,120 we stop thinking like it's us versus 389 00:08:35,120 --> 00:08:37,750 we stop thinking like it's us versus them, the conversation shifts from fear 390 00:08:37,750 --> 00:08:37,760 them, the conversation shifts from fear 391 00:08:37,760 --> 00:08:40,389 them, the conversation shifts from fear to evolution. Kershw's concept of the 392 00:08:40,389 --> 00:08:40,399 to evolution. Kershw's concept of the 393 00:08:40,399 --> 00:08:42,149 to evolution. Kershw's concept of the singularity and Neuralink's brain 394 00:08:42,149 --> 00:08:42,159 singularity and Neuralink's brain 395 00:08:42,159 --> 00:08:44,790 singularity and Neuralink's brain computer interfaces are both leading us 396 00:08:44,790 --> 00:08:44,800 computer interfaces are both leading us 397 00:08:44,800 --> 00:08:47,590 computer interfaces are both leading us toward a time where if we stay purely 398 00:08:47,590 --> 00:08:47,600 toward a time where if we stay purely 399 00:08:47,600 --> 00:08:50,829 toward a time where if we stay purely human, don't adapt and adopt these 400 00:08:50,829 --> 00:08:50,839 human, don't adapt and adopt these 401 00:08:50,839 --> 00:08:53,509 human, don't adapt and adopt these technologies, we may not stand a chance. 402 00:08:53,509 --> 00:08:53,519 technologies, we may not stand a chance. 403 00:08:53,519 --> 00:08:55,750 technologies, we may not stand a chance. We are just scratching the surface of 404 00:08:55,750 --> 00:08:55,760 We are just scratching the surface of 405 00:08:55,760 --> 00:08:58,630 We are just scratching the surface of what AI is really capable of. But if 406 00:08:58,630 --> 00:08:58,640 what AI is really capable of. But if 407 00:08:58,640 --> 00:09:00,150 what AI is really capable of. But if you're already seeing the signs and you 408 00:09:00,150 --> 00:09:00,160 you're already seeing the signs and you 409 00:09:00,160 --> 00:09:02,150 you're already seeing the signs and you want to stay ahead of this shift, tap 410 00:09:02,150 --> 00:09:02,160 want to stay ahead of this shift, tap 411 00:09:02,160 --> 00:09:04,070 want to stay ahead of this shift, tap the subscribe button below. I'll make 412 00:09:04,070 --> 00:09:04,080 the subscribe button below. I'll make 413 00:09:04,080 --> 00:09:05,670 the subscribe button below. I'll make sure that you are prepared for what 414 00:09:05,670 --> 00:09:05,680 sure that you are prepared for what 415 00:09:05,680 --> 00:09:09,509 sure that you are prepared for what comes next. For over 300,000 years, homo 416 00:09:09,509 --> 00:09:09,519 comes next. For over 300,000 years, homo 417 00:09:09,519 --> 00:09:11,829 comes next. For over 300,000 years, homo sapiens have sat at the top of the food 418 00:09:11,829 --> 00:09:11,839 sapiens have sat at the top of the food 419 00:09:11,839 --> 00:09:13,910 sapiens have sat at the top of the food chain because we the smartest. 420 00:09:13,910 --> 00:09:13,920 chain because we the smartest. 421 00:09:13,920 --> 00:09:16,230 chain because we the smartest. Intelligence was our edge. It's what 422 00:09:16,230 --> 00:09:16,240 Intelligence was our edge. It's what 423 00:09:16,240 --> 00:09:18,350 Intelligence was our edge. It's what allowed us to create fire, build 424 00:09:18,350 --> 00:09:18,360 allowed us to create fire, build 425 00:09:18,360 --> 00:09:21,110 allowed us to create fire, build civilizations, split atoms, and decode 426 00:09:21,110 --> 00:09:21,120 civilizations, split atoms, and decode 427 00:09:21,120 --> 00:09:23,829 civilizations, split atoms, and decode our own DNA. We've never had to share 428 00:09:23,829 --> 00:09:23,839 our own DNA. We've never had to share 429 00:09:23,839 --> 00:09:26,550 our own DNA. We've never had to share that throne until now. What we're 430 00:09:26,550 --> 00:09:26,560 that throne until now. What we're 431 00:09:26,560 --> 00:09:29,230 that throne until now. What we're building with AI is a new form of 432 00:09:29,230 --> 00:09:29,240 building with AI is a new form of 433 00:09:29,240 --> 00:09:32,030 building with AI is a new form of intelligence that doesn't share our 434 00:09:32,030 --> 00:09:32,040 intelligence that doesn't share our 435 00:09:32,040 --> 00:09:34,710 intelligence that doesn't share our limitations. And unlike every rival 436 00:09:34,710 --> 00:09:34,720 limitations. And unlike every rival 437 00:09:34,720 --> 00:09:37,190 limitations. And unlike every rival we've faced, this one isn't made out of 438 00:09:37,190 --> 00:09:37,200 we've faced, this one isn't made out of 439 00:09:37,200 --> 00:09:39,509 we've faced, this one isn't made out of flesh and blood. It's made of code, 440 00:09:39,509 --> 00:09:39,519 flesh and blood. It's made of code, 441 00:09:39,519 --> 00:09:42,070 flesh and blood. It's made of code, data, and exponential feedback loops. 442 00:09:42,070 --> 00:09:42,080 data, and exponential feedback loops. 443 00:09:42,080 --> 00:09:44,070 data, and exponential feedback loops. For the first time in human history, 444 00:09:44,070 --> 00:09:44,080 For the first time in human history, 445 00:09:44,080 --> 00:09:46,470 For the first time in human history, we're staring into the eyes of something 446 00:09:46,470 --> 00:09:46,480 we're staring into the eyes of something 447 00:09:46,480 --> 00:09:49,990 we're staring into the eyes of something that could surpass us permanently. And 448 00:09:49,990 --> 00:09:50,000 that could surpass us permanently. And 449 00:09:50,000 --> 00:09:52,550 that could surpass us permanently. And it's not because it hates us or wants 450 00:09:52,550 --> 00:09:52,560 it's not because it hates us or wants 451 00:09:52,560 --> 00:09:54,870 it's not because it hates us or wants power, but because we programmed it to 452 00:09:54,870 --> 00:09:54,880 power, but because we programmed it to 453 00:09:54,880 --> 00:09:57,430 power, but because we programmed it to improve and is doing exactly that. For a 454 00:09:57,430 --> 00:09:57,440 improve and is doing exactly that. For a 455 00:09:57,440 --> 00:09:59,110 improve and is doing exactly that. For a little while right now, we are still in 456 00:09:59,110 --> 00:09:59,120 little while right now, we are still in 457 00:09:59,120 --> 00:10:01,829 little while right now, we are still in charge and we can still affect how the 458 00:10:01,829 --> 00:10:01,839 charge and we can still affect how the 459 00:10:01,839 --> 00:10:04,310 charge and we can still affect how the future plays out. I'm not sure for how 460 00:10:04,310 --> 00:10:04,320 future plays out. I'm not sure for how 461 00:10:04,320 --> 00:10:05,829 future plays out. I'm not sure for how much longer though, as things are 462 00:10:05,829 --> 00:10:05,839 much longer though, as things are 463 00:10:05,839 --> 00:10:09,070 much longer though, as things are changing really fast. If this trajectory 464 00:10:09,070 --> 00:10:09,080 changing really fast. If this trajectory 465 00:10:09,080 --> 00:10:11,750 changing really fast. If this trajectory continues, AI will be the one writing 466 00:10:11,750 --> 00:10:11,760 continues, AI will be the one writing 467 00:10:11,760 --> 00:10:13,990 continues, AI will be the one writing the future, and we'll be watching it 468 00:10:13,990 --> 00:10:14,000 the future, and we'll be watching it 469 00:10:14,000 --> 00:10:17,509 the future, and we'll be watching it unfold, unable to keep up. This is the 470 00:10:17,509 --> 00:10:17,519 unfold, unable to keep up. This is the 471 00:10:17,519 --> 00:10:19,750 unfold, unable to keep up. This is the evolutionary fork in the road. So, we're 472 00:10:19,750 --> 00:10:19,760 evolutionary fork in the road. So, we're 473 00:10:19,760 --> 00:10:23,430 evolutionary fork in the road. So, we're facing a choice. Compete and lose, or 474 00:10:23,430 --> 00:10:23,440 facing a choice. Compete and lose, or 475 00:10:23,440 --> 00:10:26,470 facing a choice. Compete and lose, or evolve and coexist. If you fight the 476 00:10:26,470 --> 00:10:26,480 evolve and coexist. If you fight the 477 00:10:26,480 --> 00:10:29,350 evolve and coexist. If you fight the tide, you will be swept away. Flow with 478 00:10:29,350 --> 00:10:29,360 tide, you will be swept away. Flow with 479 00:10:29,360 --> 00:10:32,230 tide, you will be swept away. Flow with it, and we might figure out a way to 480 00:10:32,230 --> 00:10:32,240 it, and we might figure out a way to 481 00:10:32,240 --> 00:10:35,670 it, and we might figure out a way to coexist and evolve with AI while 482 00:10:35,670 --> 00:10:35,680 coexist and evolve with AI while 483 00:10:35,680 --> 00:10:38,310 coexist and evolve with AI while remaining human and who we are. This is 484 00:10:38,310 --> 00:10:38,320 remaining human and who we are. This is 485 00:10:38,320 --> 00:10:40,630 remaining human and who we are. This is a turning point for our species. Once 486 00:10:40,630 --> 00:10:40,640 a turning point for our species. Once 487 00:10:40,640 --> 00:10:43,269 a turning point for our species. Once intelligence exists outside of biology, 488 00:10:43,269 --> 00:10:43,279 intelligence exists outside of biology, 489 00:10:43,279 --> 00:10:45,670 intelligence exists outside of biology, the rules of life itself begin to 490 00:10:45,670 --> 00:10:45,680 the rules of life itself begin to 491 00:10:45,680 --> 00:10:48,150 the rules of life itself begin to change. So the real question isn't what 492 00:10:48,150 --> 00:10:48,160 change. So the real question isn't what 493 00:10:48,160 --> 00:10:51,600 change. So the real question isn't what will AI become.