1 00:00:05,200 --> 00:00:07,840 Speaker 1: What are the most important things for our students to 2 00:00:07,960 --> 00:00:10,719 Speaker 1: learn now that we're in a world where AI is 3 00:00:10,840 --> 00:00:15,160 Speaker 1: so much smarter than any human, Instead of learning skills 4 00:00:15,200 --> 00:00:18,120 Speaker 1: that may not matter as much in the future, can 5 00:00:18,160 --> 00:00:21,000 Speaker 1: AI help us to learn the important things? 6 00:00:21,320 --> 00:00:24,239 Speaker 2: And what are the important things? Do we give up on. 7 00:00:24,360 --> 00:00:28,320 Speaker 1: Education or are the very clever ways that we can 8 00:00:28,400 --> 00:00:31,160 Speaker 1: direct our teaching to get the best out of our schools. 9 00:00:31,480 --> 00:00:34,760 Speaker 1: What would Saul Khan, founder of the Kahn Academy say 10 00:00:34,800 --> 00:00:37,479 Speaker 1: about all this? Well, you're lucky because you're about to 11 00:00:37,520 --> 00:00:43,720 Speaker 1: find out. Welcome to inner Cosmos with me David Eagelman. 12 00:00:44,040 --> 00:00:47,200 Speaker 1: I'm a neuroscientist and author at Stanford and in these 13 00:00:47,240 --> 00:00:51,040 Speaker 1: episodes we dive into our three pound universe to uncover 14 00:00:51,200 --> 00:00:54,440 Speaker 1: some of the most surprising aspects of our lives and 15 00:00:54,520 --> 00:01:07,160 Speaker 1: our times. Today's episode is part two about the future 16 00:01:07,520 --> 00:01:11,399 Speaker 1: of education. In part one last week, I covered the 17 00:01:11,560 --> 00:01:17,080 Speaker 1: long road of how technology changes schooling. For example, when 18 00:01:17,080 --> 00:01:20,520 Speaker 1: the printing press was invented in fourteen forty, some thinkers 19 00:01:20,520 --> 00:01:23,800 Speaker 1: of the time were sure that invention would dumb us 20 00:01:23,800 --> 00:01:26,440 Speaker 1: all down, because now if you ask a student for 21 00:01:26,480 --> 00:01:29,440 Speaker 1: an answer, the answer is right there, It's written down. 22 00:01:29,480 --> 00:01:31,959 Speaker 1: You just pull the book off the shelf and wham, 23 00:01:32,000 --> 00:01:35,440 Speaker 1: it's as though you can pretend you know something. And 24 00:01:35,520 --> 00:01:39,959 Speaker 1: we went through similar debates when desk calculators were introduced, 25 00:01:40,200 --> 00:01:42,800 Speaker 1: will the kids become stupid? And then there was the 26 00:01:42,880 --> 00:01:45,360 Speaker 1: invention of the World Wide Web, And then when they 27 00:01:45,800 --> 00:01:48,640 Speaker 1: Google burst onto the scene, and all you had to 28 00:01:48,680 --> 00:01:52,000 Speaker 1: do was type your question into this little box and 29 00:01:52,080 --> 00:01:54,480 Speaker 1: it had indexed all the information in the world and 30 00:01:54,560 --> 00:01:57,480 Speaker 1: it could point you precisely where you needed. 31 00:01:57,200 --> 00:01:58,760 Speaker 2: To go to find the answer. 32 00:01:59,400 --> 00:02:03,120 Speaker 1: So suddenly it didn't matter what question the teacher asked you. 33 00:02:03,520 --> 00:02:05,440 Speaker 2: As long as somebody somewhere. 34 00:02:05,080 --> 00:02:07,240 Speaker 1: In the world knew the answer and had typed it 35 00:02:07,280 --> 00:02:09,919 Speaker 1: on the web page, it was just one click away 36 00:02:09,919 --> 00:02:13,920 Speaker 1: from you. In all these cases, educators were forced to 37 00:02:14,120 --> 00:02:18,840 Speaker 1: change the way they asked questions and assessed a student's knowledge. 38 00:02:19,200 --> 00:02:21,080 Speaker 2: And you know what, it wasn't that hard. 39 00:02:21,880 --> 00:02:24,880 Speaker 1: So where are we with technology in general and AI 40 00:02:24,960 --> 00:02:31,280 Speaker 1: in particular right now? I made three main arguments in 41 00:02:31,360 --> 00:02:36,160 Speaker 1: last week's episode. I argued that the extant technology wires 42 00:02:36,240 --> 00:02:39,880 Speaker 1: the brain of each generation differently, as in how they're 43 00:02:39,919 --> 00:02:42,960 Speaker 1: wired to best learn and the ways they seek information. 44 00:02:43,520 --> 00:02:46,400 Speaker 1: Then I gave many examples of the ways in which 45 00:02:46,440 --> 00:02:50,239 Speaker 1: AI can be used to massively improve the way we teach, 46 00:02:50,680 --> 00:02:54,600 Speaker 1: from individualized tutoring to reducing the workload on the teachers, 47 00:02:54,840 --> 00:02:59,400 Speaker 1: to training critical thinking and magnifying creativity. If you want 48 00:02:59,440 --> 00:03:02,559 Speaker 1: to hear more about that, please check out that last episode. 49 00:03:02,639 --> 00:03:05,440 Speaker 1: And finally, I made the argument that the purpose for 50 00:03:05,560 --> 00:03:10,960 Speaker 1: schooling for education is to give the students roots and wings, 51 00:03:11,320 --> 00:03:14,600 Speaker 1: as Gerta phrased it, which just means giving them the 52 00:03:14,680 --> 00:03:18,480 Speaker 1: foundations that they need for critical thinking and the ability 53 00:03:18,520 --> 00:03:21,880 Speaker 1: to create new things and go off in novel directions. 54 00:03:22,760 --> 00:03:23,760 Speaker 2: So today we're going to. 55 00:03:23,680 --> 00:03:25,800 Speaker 1: Get someone else's take on the state of things, and 56 00:03:25,840 --> 00:03:29,600 Speaker 1: not just anyone, but someone who has single handedly had 57 00:03:29,639 --> 00:03:33,320 Speaker 1: perhaps the largest impact on the planet's education. 58 00:03:33,440 --> 00:03:35,640 Speaker 2: That I know. That's all Con. 59 00:03:35,920 --> 00:03:37,640 Speaker 1: And if you don't know his name, you surely know 60 00:03:37,720 --> 00:03:42,280 Speaker 1: his creation, the Con Academy, which is an educational platform 61 00:03:42,640 --> 00:03:47,280 Speaker 1: that provides free and very high quality online courses. It 62 00:03:47,320 --> 00:03:50,760 Speaker 1: does this across math and science, and history and economics. 63 00:03:50,440 --> 00:03:51,680 Speaker 2: Anything you want to learn. 64 00:03:51,840 --> 00:03:56,440 Speaker 1: It has video lessons and interactive exercises and learning tools, 65 00:03:56,840 --> 00:03:59,760 Speaker 1: and the key is, unlike a normal school system, this 66 00:04:00,080 --> 00:04:04,000 Speaker 1: allows users to progress at their own pace, so Saul 67 00:04:04,080 --> 00:04:07,280 Speaker 1: has a unique vantage point to view the world's education 68 00:04:07,840 --> 00:04:10,520 Speaker 1: and as my neighbor here in Silicon Valley, he has 69 00:04:10,560 --> 00:04:14,600 Speaker 1: the opportunity to see the latest greatest tech and understand 70 00:04:14,640 --> 00:04:16,120 Speaker 1: where things are going. 71 00:04:16,720 --> 00:04:20,599 Speaker 2: So, without further ado, here's my interview with Saul con. 72 00:04:24,800 --> 00:04:28,640 Speaker 1: Okay, sal So, you're famous for having started the con Academy. 73 00:04:28,720 --> 00:04:30,160 Speaker 2: Tell us about the history of that. 74 00:04:30,640 --> 00:04:32,480 Speaker 3: Oh hi, you know, I guess you go back about 75 00:04:32,520 --> 00:04:35,560 Speaker 3: twenty years ago. A little over twenty years ago. My 76 00:04:35,640 --> 00:04:38,080 Speaker 3: original background was in technology, and I go to business 77 00:04:38,080 --> 00:04:39,720 Speaker 3: school and I find myself. I'm working at a hedge 78 00:04:39,720 --> 00:04:42,000 Speaker 3: fund in Boston. Two thousand and four. I'm a year 79 00:04:42,000 --> 00:04:45,040 Speaker 3: out of business school. I get married. My wife grew 80 00:04:45,120 --> 00:04:47,080 Speaker 3: up in New Jersey, so all of my family from 81 00:04:47,120 --> 00:04:49,320 Speaker 3: New Orleans, where I was born and raised, comes to 82 00:04:49,360 --> 00:04:51,839 Speaker 3: New Jersey for the wedding. While they were staying with me. 83 00:04:52,520 --> 00:04:54,599 Speaker 3: My cousin, it just came out of conversation was having 84 00:04:54,600 --> 00:04:57,080 Speaker 3: trouble with math Navia. She was twelve years old, and 85 00:04:57,120 --> 00:04:58,840 Speaker 3: when I asked her about it is unit conversion, etc. 86 00:04:59,040 --> 00:05:00,839 Speaker 3: I told Navy, I'm a hundred percent sure you can. 87 00:05:01,040 --> 00:05:03,240 Speaker 3: You can learn this material if you're up for it. 88 00:05:03,279 --> 00:05:05,040 Speaker 3: I'm happy to tutor you remotely when you go back 89 00:05:05,080 --> 00:05:07,920 Speaker 3: to New Orleans. She agrees, she goes back. We start 90 00:05:07,960 --> 00:05:11,080 Speaker 3: working together, slowly, but surely she gets unit conversion. She 91 00:05:11,080 --> 00:05:12,960 Speaker 3: gets caught up with her class, littlehead of her class. 92 00:05:13,400 --> 00:05:15,560 Speaker 3: Then I'd become what I call a tiger cousin. And 93 00:05:15,640 --> 00:05:19,440 Speaker 3: I call up her school and they. 94 00:05:19,320 --> 00:05:21,000 Speaker 2: Said, who are you? I said, I'm her cousin. 95 00:05:21,480 --> 00:05:24,320 Speaker 3: And I actually convinced them to let her retake a 96 00:05:24,360 --> 00:05:26,280 Speaker 3: placement exam because I thought, like, well, she knows her 97 00:05:26,279 --> 00:05:29,600 Speaker 3: material now, and they did, and she went from being 98 00:05:29,680 --> 00:05:31,560 Speaker 3: like a for lack of better word, to remut, like 99 00:05:31,800 --> 00:05:34,440 Speaker 3: in a slower track, to not being in a faster track. 100 00:05:34,520 --> 00:05:35,840 Speaker 2: So I was like, well, this is cool. Like I'm 101 00:05:35,880 --> 00:05:37,200 Speaker 2: able to connect with a young cousin. 102 00:05:38,000 --> 00:05:40,359 Speaker 3: I start tutoring her younger brother's words preadient, the family 103 00:05:40,400 --> 00:05:42,680 Speaker 3: free tutoring is going on. Before I know it, I'm 104 00:05:42,720 --> 00:05:45,440 Speaker 3: tutoring ten to fifteen cousins, family friends. You know, I 105 00:05:45,440 --> 00:05:49,080 Speaker 3: saw a common pattern that they were struggling, not because 106 00:05:49,080 --> 00:05:51,040 Speaker 3: they weren't going to good schools, not because they didn't 107 00:05:51,040 --> 00:05:53,719 Speaker 3: have good teachers, not because they weren't hard working or bright. 108 00:05:54,040 --> 00:05:56,159 Speaker 3: It's because they had gaps in their learning, especially in 109 00:05:56,160 --> 00:05:57,920 Speaker 3: a topic like math. But even you know, part of 110 00:05:57,960 --> 00:05:59,880 Speaker 3: the reason we were having trouble in science was gaps 111 00:06:00,040 --> 00:06:02,840 Speaker 3: sometimes in their math. So I started writing software for 112 00:06:02,880 --> 00:06:06,680 Speaker 3: them to kind of give them practice them being your cousins, 113 00:06:06,680 --> 00:06:08,760 Speaker 3: my cousins. Yeah, this was back in the day. If 114 00:06:08,800 --> 00:06:12,599 Speaker 3: you had to look for practice problems on factoring polynomials, 115 00:06:12,880 --> 00:06:14,640 Speaker 3: there were stuff on the internet, but they weren't very 116 00:06:14,680 --> 00:06:16,080 Speaker 3: good and they were very inconsistent. It's like, well, I 117 00:06:16,080 --> 00:06:17,680 Speaker 3: think I could write a little piece of software that'll 118 00:06:17,880 --> 00:06:20,640 Speaker 3: generate problems for them, and then I want to keep 119 00:06:20,680 --> 00:06:22,680 Speaker 3: track of how they're doing. And then that was the 120 00:06:22,720 --> 00:06:25,120 Speaker 3: first conducaut I mean, nothing to do with videos. Then 121 00:06:25,480 --> 00:06:26,400 Speaker 3: this was two thousand and five. 122 00:06:26,400 --> 00:06:27,680 Speaker 2: Then how many people were using it? 123 00:06:28,279 --> 00:06:31,039 Speaker 3: Initially just my cousins, and I was seeing benefit and 124 00:06:31,040 --> 00:06:33,600 Speaker 3: I could keeping track of them. Yeah, ten to fifteen cousins, 125 00:06:34,520 --> 00:06:37,280 Speaker 3: and then words started to spread. Their friends started to 126 00:06:37,279 --> 00:06:38,680 Speaker 3: sign up to it because they're like, oh, this is 127 00:06:38,680 --> 00:06:41,440 Speaker 3: really helping me get good at math, and so their 128 00:06:41,480 --> 00:06:43,200 Speaker 3: friends and at some point had to shut it down. 129 00:06:43,200 --> 00:06:44,920 Speaker 3: There got to about ten thousand folks who were trying 130 00:06:44,920 --> 00:06:47,000 Speaker 3: to use just the software, and it was taking down 131 00:06:47,040 --> 00:06:50,280 Speaker 3: my thirty dollars a month web hosting, and that wasn't 132 00:06:50,320 --> 00:06:53,360 Speaker 3: engineered perfectly either to scale to that. So I actually 133 00:06:53,360 --> 00:06:55,159 Speaker 3: just didn't allow anyone else to sign up, which is 134 00:06:55,160 --> 00:06:57,920 Speaker 3: like the opposite of Silicon Valley, Like, no, I don't 135 00:06:58,000 --> 00:07:00,080 Speaker 3: like it, there's too many people here. And then and 136 00:07:00,640 --> 00:07:02,800 Speaker 3: I was showing this software off at a dinner party. 137 00:07:02,800 --> 00:07:04,880 Speaker 3: And by this point we had moved out to northern California, 138 00:07:05,360 --> 00:07:07,120 Speaker 3: and I was at a dinner party in San Mateo 139 00:07:07,600 --> 00:07:10,080 Speaker 3: and the host and I was showing off this stuff 140 00:07:10,120 --> 00:07:11,600 Speaker 3: I was making, and she said, well, how are you 141 00:07:11,640 --> 00:07:14,040 Speaker 3: scaling your lessons. I'm like, I'm not. He's like, whight 142 00:07:14,040 --> 00:07:16,480 Speaker 3: you record some videos on YouTube for your family. I said, 143 00:07:16,480 --> 00:07:19,280 Speaker 3: that's a dumb idea. You know that's cat's playing piano. 144 00:07:19,400 --> 00:07:22,720 Speaker 3: That's not serious. But I did it anyway. Two thousand 145 00:07:22,720 --> 00:07:26,720 Speaker 3: and six, Okay, and yeah, I just started making mini lessons. 146 00:07:26,720 --> 00:07:28,560 Speaker 3: Actually YouTube forced them to be mini lessons at a 147 00:07:28,560 --> 00:07:30,560 Speaker 3: ten minute limit at the time, on things that I 148 00:07:30,600 --> 00:07:32,600 Speaker 3: found myself having to re explain to my cousins a lot, 149 00:07:32,640 --> 00:07:34,520 Speaker 3: Like I'd explain least common multiple to one of my 150 00:07:34,560 --> 00:07:36,840 Speaker 3: cousins one week and then the next week, I'm re 151 00:07:36,880 --> 00:07:39,440 Speaker 3: explaining it to another cousin, and so I said, well, okay, 152 00:07:39,520 --> 00:07:41,400 Speaker 3: I know least common multiples shows up a lot, you know, 153 00:07:41,400 --> 00:07:42,120 Speaker 3: negative exposed. 154 00:07:42,120 --> 00:07:42,920 Speaker 2: So I just kept doing it. 155 00:07:42,960 --> 00:07:45,800 Speaker 3: I started emailing them and saying, any question you have, 156 00:07:45,920 --> 00:07:48,560 Speaker 3: I'll do it asynchronously. You email me the question, I'll 157 00:07:48,560 --> 00:07:50,000 Speaker 3: give you a video explanation of it. 158 00:07:50,240 --> 00:07:50,680 Speaker 2: Wow. 159 00:07:50,720 --> 00:07:53,080 Speaker 3: And so we started building up a library of that. 160 00:07:53,200 --> 00:07:56,560 Speaker 3: And you know, my cousins really appreciated it. I sometimes joke, 161 00:07:56,640 --> 00:07:58,280 Speaker 3: but it's true. They said they liked me better on 162 00:07:58,320 --> 00:08:02,000 Speaker 3: YouTube than in person. And what they were saying is 163 00:08:02,040 --> 00:08:05,040 Speaker 3: that they like the asynchronous, the on demand, the no shame. 164 00:08:05,440 --> 00:08:06,800 Speaker 3: But we still were getting on the phone and they 165 00:08:06,800 --> 00:08:10,600 Speaker 3: did appreciate me as a human being being in their life, 166 00:08:11,360 --> 00:08:12,880 Speaker 3: and we were able to go deeper on the phone. 167 00:08:13,040 --> 00:08:15,440 Speaker 3: But and yeah, and then one thing led to another. 168 00:08:15,520 --> 00:08:17,600 Speaker 3: Two thousand and seven, two thousand and eight, fifteen to 169 00:08:17,640 --> 00:08:18,480 Speaker 3: one hundred thousand folks. 170 00:08:18,760 --> 00:08:19,680 Speaker 2: I was still at the Hedge Fund. 171 00:08:20,080 --> 00:08:21,960 Speaker 3: I set up as a nonprofit in two thousand and eight, 172 00:08:22,040 --> 00:08:24,520 Speaker 3: the mission Free World Class Education for Anyone, Anywhere. Two 173 00:08:24,560 --> 00:08:26,800 Speaker 3: thousand and nine, quit the day job. Two thousand and 174 00:08:26,840 --> 00:08:28,720 Speaker 3: nine was a hard year, but twenty ten, we got 175 00:08:28,720 --> 00:08:30,680 Speaker 3: our first real funding to become an organization. 176 00:08:30,880 --> 00:08:31,160 Speaker 2: Wow. 177 00:08:31,200 --> 00:08:34,280 Speaker 1: And how many people use con Academy now, Oh, you know, 178 00:08:34,280 --> 00:08:36,760 Speaker 1: it depends on how you account for it. I think 179 00:08:36,800 --> 00:08:39,719 Speaker 1: we're one hundred and seventy million registered users, but. 180 00:08:39,760 --> 00:08:41,520 Speaker 2: You know, not everyone uses it all the time. 181 00:08:42,880 --> 00:08:44,520 Speaker 3: You know, in a given year, it's probably closer to 182 00:08:44,520 --> 00:08:46,320 Speaker 3: about one hundred million folks are using it. 183 00:08:46,440 --> 00:08:46,960 Speaker 2: Incredible. 184 00:08:47,000 --> 00:08:49,960 Speaker 1: I mean, that must make you the biggest single educator 185 00:08:50,520 --> 00:08:51,719 Speaker 1: and now I know you have a whole team of 186 00:08:51,760 --> 00:08:54,960 Speaker 1: people doing it now, but the biggest educational organization in 187 00:08:55,000 --> 00:08:55,280 Speaker 1: the world. 188 00:08:55,320 --> 00:08:57,920 Speaker 2: I would assume. Yes, it depends how you account for it. 189 00:08:58,160 --> 00:09:00,079 Speaker 3: Yeah, I mean, but in some ways yeah, yeah, I 190 00:09:00,120 --> 00:09:01,960 Speaker 3: mean some people can make an argument that you know, 191 00:09:02,040 --> 00:09:06,079 Speaker 3: YouTube amongst other things, does some stuff. But yeah, I look, 192 00:09:06,120 --> 00:09:07,959 Speaker 3: but I always point out to the team that's nice. 193 00:09:08,000 --> 00:09:09,880 Speaker 3: But we have a long way to go because even 194 00:09:10,080 --> 00:09:13,560 Speaker 3: of that one hundred million, there's probably about three million 195 00:09:13,640 --> 00:09:15,640 Speaker 3: that are like really getting a lot out of it. 196 00:09:15,720 --> 00:09:18,640 Speaker 3: They're using it almost like you know, regularly, and those 197 00:09:18,679 --> 00:09:20,040 Speaker 3: people are really getting an education. 198 00:09:20,280 --> 00:09:20,920 Speaker 2: Incredible. 199 00:09:21,120 --> 00:09:23,600 Speaker 1: Okay, So now we find ourselves in twenty twenty five 200 00:09:23,960 --> 00:09:26,640 Speaker 1: and AI is on the scene, and so how are 201 00:09:26,679 --> 00:09:29,680 Speaker 1: you using AI in the con academy. 202 00:09:29,800 --> 00:09:32,120 Speaker 3: Yeah, And I remind my because it's very easy to 203 00:09:32,120 --> 00:09:35,280 Speaker 3: get nambored with technology, especially things as cool as generative AI. 204 00:09:35,760 --> 00:09:38,840 Speaker 3: We had a preview if you rewind up two two 205 00:09:38,880 --> 00:09:42,000 Speaker 3: and a half years summer of twenty twenty two. Sam 206 00:09:42,000 --> 00:09:44,600 Speaker 3: Altman and Greg Brockman reach out to me at the 207 00:09:44,600 --> 00:09:48,439 Speaker 3: time and it's just this kind of cryptic email saying 208 00:09:48,520 --> 00:09:50,040 Speaker 3: and I didn't know them that well. I'd run into 209 00:09:50,080 --> 00:09:52,600 Speaker 3: them at some conferences and this and that, and they said, Hey, 210 00:09:52,800 --> 00:09:55,040 Speaker 3: we're training our new model. We think this is the 211 00:09:55,080 --> 00:09:57,559 Speaker 3: one that's going to really get people interested. We want 212 00:09:57,559 --> 00:10:00,000 Speaker 3: to launch with some social positive use cases that you 213 00:10:00,000 --> 00:10:02,800 Speaker 3: GPT four it ended up being GPT four And this 214 00:10:02,920 --> 00:10:06,920 Speaker 3: was this was about five months before chat GPT came out, 215 00:10:07,240 --> 00:10:09,440 Speaker 3: which actually was a bit of an accident. Chat which 216 00:10:09,440 --> 00:10:11,440 Speaker 3: and Chat GPT wasn't even built on GPT four. It 217 00:10:11,440 --> 00:10:14,240 Speaker 3: was built on GPT three point five. But when they 218 00:10:14,880 --> 00:10:16,880 Speaker 3: emailed me, I was like, I was curious. I had 219 00:10:16,880 --> 00:10:19,120 Speaker 3: seen GPT two and GPT three. I thought they were 220 00:10:19,200 --> 00:10:21,560 Speaker 3: very cool, but I didn't think there were anywhere close 221 00:10:21,840 --> 00:10:23,040 Speaker 3: to ready for prime time. 222 00:10:23,240 --> 00:10:24,640 Speaker 2: But as a nerd, I just wanted to see what 223 00:10:24,679 --> 00:10:27,360 Speaker 2: they were going to do. So two weeks later they 224 00:10:27,400 --> 00:10:30,640 Speaker 2: had finished the training and they had myself and our 225 00:10:30,679 --> 00:10:34,000 Speaker 2: chief Learning officer, Kristin. We were on a zoom with 226 00:10:34,040 --> 00:10:36,599 Speaker 2: them and they had a little chat interface and this 227 00:10:36,880 --> 00:10:37,280 Speaker 2: was new too. 228 00:10:37,320 --> 00:10:39,199 Speaker 3: People don't remember these, you know, just even have a 229 00:10:39,280 --> 00:10:41,640 Speaker 3: chat interface with an AI. And they had an ap 230 00:10:41,760 --> 00:10:45,240 Speaker 3: biology question and there's actually context why they picked ap biology. 231 00:10:45,559 --> 00:10:47,719 Speaker 3: But they said, so, what's the answer and I said, oh, 232 00:10:48,400 --> 00:10:51,680 Speaker 3: c it's osmosis. And then they clicked enter and the 233 00:10:51,760 --> 00:10:54,320 Speaker 3: AI said, oh the answer ce. I'm like, okay, ask 234 00:10:54,360 --> 00:10:57,200 Speaker 3: it why. It explained it. And people are used to 235 00:10:57,200 --> 00:10:59,760 Speaker 3: this now, but in the summer of twenty twenty two, 236 00:11:00,360 --> 00:11:03,480 Speaker 3: this was different than anything anyone had ever seen before. 237 00:11:03,480 --> 00:11:06,160 Speaker 3: And I'm like, okay, this is weird. Ask it why 238 00:11:06,200 --> 00:11:09,520 Speaker 3: the other choices are incorrect? It explained it beautifully, I said, 239 00:11:09,800 --> 00:11:11,280 Speaker 3: ask it tried another question like that. 240 00:11:11,800 --> 00:11:14,080 Speaker 2: It did a great job. And that's when I was like, oh, 241 00:11:14,120 --> 00:11:15,680 Speaker 2: what is this like? Am I being punked? 242 00:11:15,760 --> 00:11:15,800 Speaker 1: Like? 243 00:11:16,800 --> 00:11:18,920 Speaker 3: But we saw it before, not all of open AI 244 00:11:19,040 --> 00:11:22,240 Speaker 3: even saw it. But then they gave Kristen myself access 245 00:11:22,280 --> 00:11:25,080 Speaker 3: to it over the weekend and we couldn't sleep, and 246 00:11:25,120 --> 00:11:27,280 Speaker 3: we couldn't tell anyone about it, Like, because this was 247 00:11:27,280 --> 00:11:29,480 Speaker 3: the biggest secret, I couldn't tell my wife about it. 248 00:11:30,720 --> 00:11:32,520 Speaker 3: But I was like middle of the night, I was 249 00:11:32,520 --> 00:11:34,000 Speaker 3: slacking christ and I was like, look, I just got 250 00:11:34,080 --> 00:11:35,760 Speaker 3: it to write the Decoration of Independence and the voice 251 00:11:35,760 --> 00:11:36,360 Speaker 3: of Donald Trump. 252 00:11:36,360 --> 00:11:37,079 Speaker 2: Isn't this hilarious? 253 00:11:37,080 --> 00:11:39,160 Speaker 3: And I'm writing poems and I'm doing this and look, 254 00:11:39,160 --> 00:11:40,520 Speaker 3: if you prompt it the right way, can. 255 00:11:40,400 --> 00:11:41,160 Speaker 2: Act as a tutor. 256 00:11:41,679 --> 00:11:45,120 Speaker 3: And that's when going back to what Kin Academy was 257 00:11:45,120 --> 00:11:47,959 Speaker 3: always about. As I mentioned, I started tutoring my cousins 258 00:11:48,240 --> 00:11:50,680 Speaker 3: and everything that we've done over the last fifteen twenty years, 259 00:11:50,679 --> 00:11:53,040 Speaker 3: it's really about how do we use technology to approximate 260 00:11:53,040 --> 00:11:55,840 Speaker 3: elements of a tutor on demand bite sized video kind 261 00:11:55,880 --> 00:11:59,000 Speaker 3: of approximates an on demand explanation from a tutor. These 262 00:11:59,000 --> 00:12:02,600 Speaker 3: practice problems that you're doing, the problems that you need 263 00:12:02,640 --> 00:12:04,360 Speaker 3: to do as opposed to what everyone else is doing. 264 00:12:04,679 --> 00:12:06,000 Speaker 2: That's what you would do with a little bit of 265 00:12:06,080 --> 00:12:07,760 Speaker 2: a good tutor. They would give you what you need. 266 00:12:08,280 --> 00:12:11,720 Speaker 3: And when we realize that you could steer or prompt 267 00:12:11,920 --> 00:12:15,800 Speaker 3: these this new generation of models to really be not 268 00:12:15,880 --> 00:12:18,000 Speaker 3: just pretend to be the tutor, but really have strong 269 00:12:18,120 --> 00:12:19,520 Speaker 3: tutor moves, so to speak. 270 00:12:19,960 --> 00:12:21,440 Speaker 2: We're like, Okay, this thing has. 271 00:12:21,280 --> 00:12:26,600 Speaker 3: Issues, It hallucinates, it makes horrible math errors. But if 272 00:12:26,600 --> 00:12:28,599 Speaker 3: we assume that those things can either be mitigated or 273 00:12:28,640 --> 00:12:31,000 Speaker 3: they're going to get a lot better, which has happened. 274 00:12:31,360 --> 00:12:32,920 Speaker 3: This is going to be a game changer for us 275 00:12:32,960 --> 00:12:34,600 Speaker 3: being able to scale up what we think is world 276 00:12:34,600 --> 00:12:38,720 Speaker 3: class education, which is more personalization. Giving what Navia had 277 00:12:38,760 --> 00:12:41,600 Speaker 3: in sal what if every student we could eventually get there. 278 00:12:41,920 --> 00:12:44,240 Speaker 1: So let's double click on that idea about an AI tutor, 279 00:12:44,280 --> 00:12:46,880 Speaker 1: because I heard it talk from Steve Jobs way back 280 00:12:46,920 --> 00:12:48,640 Speaker 1: in the day, maybe in the nineties or something. He said, 281 00:12:48,720 --> 00:12:51,560 Speaker 1: just imagine if everyone had Aristotle as a tutor, if 282 00:12:51,600 --> 00:12:54,680 Speaker 1: I'm remembering what he said, treg But I asked my 283 00:12:54,840 --> 00:12:56,839 Speaker 1: son about that at someonhere. He's thirteen years old, and 284 00:12:56,920 --> 00:12:58,559 Speaker 1: I said, you know, hey, what would it be like 285 00:12:58,600 --> 00:13:00,319 Speaker 1: for you to have Aristotles a tutor? And he's said, 286 00:13:00,600 --> 00:13:02,680 Speaker 1: I probably wouldn't do much with it. And the reason 287 00:13:02,720 --> 00:13:06,120 Speaker 1: he said that is because, first of all, his internal 288 00:13:06,200 --> 00:13:09,160 Speaker 1: model of what he knows and what Aristotle knows from 289 00:13:09,200 --> 00:13:11,520 Speaker 1: all his writings they're different. So you wouldn't even know 290 00:13:11,559 --> 00:13:13,680 Speaker 1: what to ask him. But besides that, what he wants 291 00:13:13,720 --> 00:13:15,840 Speaker 1: to do is be with friends and play video games stuff. 292 00:13:16,200 --> 00:13:19,080 Speaker 1: So that model of tutor, I think isn't going to work. 293 00:13:19,120 --> 00:13:22,360 Speaker 1: But what you're doing is different. It's not just someone 294 00:13:22,400 --> 00:13:24,800 Speaker 1: available to answer questions for you. So tell us about 295 00:13:24,800 --> 00:13:29,559 Speaker 1: how you think about surfacing information in the learner. 296 00:13:30,000 --> 00:13:31,600 Speaker 3: No, that's right, and I want to be clear where 297 00:13:31,600 --> 00:13:33,640 Speaker 3: you know, it's a journey and it's only going to 298 00:13:33,679 --> 00:13:36,000 Speaker 3: get hopefully better over time. But immediately when we started 299 00:13:36,000 --> 00:13:37,840 Speaker 3: seeing all this, we said, all right, well, some of 300 00:13:37,840 --> 00:13:40,480 Speaker 3: the limitations of classic con Academy is you could watch 301 00:13:40,520 --> 00:13:42,439 Speaker 3: a five minute video, but what if you have a question. 302 00:13:42,840 --> 00:13:44,480 Speaker 3: If you have a peer around, you could ask them. 303 00:13:44,520 --> 00:13:45,760 Speaker 3: If you have a teacher around, you can ask them. 304 00:13:45,800 --> 00:13:48,120 Speaker 3: But what if you're a young girl in Afghanistan and 305 00:13:48,480 --> 00:13:49,559 Speaker 3: you don't have a school. 306 00:13:49,880 --> 00:13:50,560 Speaker 2: What do you do? Then? 307 00:13:50,880 --> 00:13:53,880 Speaker 3: If you're working on a on a exercise and you 308 00:13:54,000 --> 00:13:55,760 Speaker 3: just got stuck. We try to make it as gradual 309 00:13:55,800 --> 00:13:57,760 Speaker 3: as possible, but sometimes you just get stuck, and that's 310 00:13:57,800 --> 00:14:00,240 Speaker 3: an that's the most likely time that you'll just engage. 311 00:14:00,800 --> 00:14:02,040 Speaker 2: What if you could get unstuck? 312 00:14:02,280 --> 00:14:05,200 Speaker 3: So the first iteration of what we called what we 313 00:14:05,240 --> 00:14:08,160 Speaker 3: still call Conmigo, which is the AI tutor on con Academy, 314 00:14:08,880 --> 00:14:11,400 Speaker 3: was supporting those use cases. Now, to your point and 315 00:14:11,440 --> 00:14:14,720 Speaker 3: to your son's point, what we see is about ten 316 00:14:14,840 --> 00:14:17,839 Speaker 3: fifteen percent. We haven't, you know, had a formal study 317 00:14:17,880 --> 00:14:19,920 Speaker 3: of this, but what we're observing about ten fifteen percent 318 00:14:19,920 --> 00:14:22,600 Speaker 3: of students make good use of that. They're the students 319 00:14:22,600 --> 00:14:24,720 Speaker 3: that if they watch a video, they're still curious. They're like, 320 00:14:24,720 --> 00:14:26,440 Speaker 3: what about this, or how about this? What if we 321 00:14:26,520 --> 00:14:29,920 Speaker 3: change this variable? And so they know how to ask that. 322 00:14:30,240 --> 00:14:32,560 Speaker 3: But there's another eighty percent of students who are like, oh, 323 00:14:32,640 --> 00:14:35,080 Speaker 3: you know whatever, this isn't my thing. And so what 324 00:14:35,120 --> 00:14:38,920 Speaker 3: we're realizing is to have the benefits of an Aristotle, 325 00:14:39,000 --> 00:14:41,920 Speaker 3: and Aristotle was Alexander the great tutor. Aristotle didn't just 326 00:14:41,920 --> 00:14:43,800 Speaker 3: sit there and wait, I'm guessing, didn't just sit there 327 00:14:43,840 --> 00:14:46,840 Speaker 3: and wait for Alexander to ask him another question. He said, Okay, 328 00:14:46,840 --> 00:14:49,080 Speaker 3: this is what we're going to do, young Alexander. And 329 00:14:49,360 --> 00:14:50,920 Speaker 3: by the way, how come you didn't do your homework. 330 00:14:51,120 --> 00:14:53,200 Speaker 3: I'm going to go talk to was it Prince phil 331 00:14:53,360 --> 00:14:55,720 Speaker 3: King Philip or verbal Macedonia? 332 00:14:55,760 --> 00:14:57,040 Speaker 2: Yeah, his dad, right. 333 00:14:57,200 --> 00:14:59,360 Speaker 3: Philip of Macedonia I'm going to go talk to King 334 00:14:59,360 --> 00:15:01,840 Speaker 3: Philip about you're not doing your work. 335 00:15:02,080 --> 00:15:03,640 Speaker 2: And that's what I used to do with my cousins. 336 00:15:03,880 --> 00:15:06,160 Speaker 3: Yes, I would answer their questions, but most of the 337 00:15:06,160 --> 00:15:07,920 Speaker 3: time I was driving, I'd be like, this is what 338 00:15:07,960 --> 00:15:10,320 Speaker 3: we're going to do today because I noticed this, And 339 00:15:10,400 --> 00:15:12,920 Speaker 3: if they showed up unprepared, I would hold them accountable. 340 00:15:12,920 --> 00:15:14,480 Speaker 3: And sometimes I would call their mom or their dad 341 00:15:14,560 --> 00:15:16,800 Speaker 3: up and say, look, you got to make sure they 342 00:15:16,840 --> 00:15:19,040 Speaker 3: have time for this. And so we are working on 343 00:15:19,120 --> 00:15:20,960 Speaker 3: ways that the AI can be more proactive. 344 00:15:21,400 --> 00:15:21,600 Speaker 2: Now. 345 00:15:21,600 --> 00:15:24,080 Speaker 3: At the end of the day, we do think that 346 00:15:23,760 --> 00:15:27,040 Speaker 3: that teacher and if a student has access to a teacher, 347 00:15:27,640 --> 00:15:29,320 Speaker 3: which is true of most of the world we live in, 348 00:15:29,360 --> 00:15:30,960 Speaker 3: there's parts of the world where they don't, But if 349 00:15:31,000 --> 00:15:33,320 Speaker 3: they have access to a teacher, the teacher is still 350 00:15:33,360 --> 00:15:37,920 Speaker 3: the biggest motivational mechanic for students. And so we are 351 00:15:37,960 --> 00:15:40,360 Speaker 3: thinking about how can the AI both support the teacher 352 00:15:40,400 --> 00:15:42,840 Speaker 3: as a teaching assistant helping them with some of their tasks, 353 00:15:43,280 --> 00:15:46,680 Speaker 3: but then also be a connective tissue and the teacher 354 00:15:46,680 --> 00:15:49,480 Speaker 3: can hold the student accountable on working with the A 355 00:15:49,520 --> 00:15:51,040 Speaker 3: and then the a I can report back to the 356 00:15:51,120 --> 00:15:54,360 Speaker 3: to the teacher. So it's a journey on how we 357 00:15:54,400 --> 00:15:56,920 Speaker 3: can make it more and more part of that loop 358 00:15:57,120 --> 00:16:00,000 Speaker 3: and it can drive more and more, but once again 359 00:16:00,360 --> 00:16:03,320 Speaker 3: not replacing the humans, but but optimizing the connection with. 360 00:16:03,320 --> 00:16:04,960 Speaker 2: The humans so that they don't have to worry about 361 00:16:05,000 --> 00:16:07,000 Speaker 2: some of the busy work. Great. 362 00:16:07,040 --> 00:16:08,560 Speaker 1: So I want to come back to that point about 363 00:16:08,560 --> 00:16:11,080 Speaker 1: what this will mean for teachers, but I want to 364 00:16:11,160 --> 00:16:13,600 Speaker 1: quickly interject something, which is so I'm sort of a 365 00:16:13,720 --> 00:16:16,320 Speaker 1: history buff, but I blanked on this issue about King 366 00:16:16,320 --> 00:16:19,800 Speaker 1: Philip of Macedonia and so, but my question is, I 367 00:16:19,800 --> 00:16:23,040 Speaker 1: mean the question everyone's asking currently is, you know, with 368 00:16:23,200 --> 00:16:26,720 Speaker 1: AI in the world and that's not going away, what 369 00:16:26,880 --> 00:16:29,920 Speaker 1: is the importance of memorizing facts? 370 00:16:29,960 --> 00:16:31,600 Speaker 2: How will education change? 371 00:16:32,040 --> 00:16:34,800 Speaker 1: Will it be important for children to learn a lot, 372 00:16:34,840 --> 00:16:36,920 Speaker 1: you know, the Battle of Hastings was ten sixty six, 373 00:16:37,280 --> 00:16:38,920 Speaker 1: or will it not be so important because they can 374 00:16:38,960 --> 00:16:40,640 Speaker 1: retrieve that information to a point? 375 00:16:40,720 --> 00:16:41,680 Speaker 2: What's your take on that. 376 00:16:42,360 --> 00:16:45,160 Speaker 3: I'm solidly in the camp that content knowledge is really 377 00:16:45,200 --> 00:16:49,640 Speaker 3: valuable because and we've seen this pattern before Calculator comes 378 00:16:49,640 --> 00:16:51,040 Speaker 3: out in the sixties and seventies, We're like. 379 00:16:51,000 --> 00:16:52,760 Speaker 2: Oh, no, one has to learned arithmetic anymore. Well, it 380 00:16:52,760 --> 00:16:53,120 Speaker 2: turns out. 381 00:16:53,120 --> 00:16:55,040 Speaker 3: The people who benefit the most from calculators are the 382 00:16:55,040 --> 00:16:58,680 Speaker 3: people who understood arithmetic and why why They know how 383 00:16:58,720 --> 00:16:59,960 Speaker 3: to use the tool. They know when the tool can 384 00:17:00,200 --> 00:17:01,760 Speaker 3: be used, they know when the tool can't be used. 385 00:17:01,800 --> 00:17:04,639 Speaker 3: They know when there's a precision error in your calculator. 386 00:17:05,040 --> 00:17:06,840 Speaker 3: You know, they know that one third is the same 387 00:17:06,840 --> 00:17:09,439 Speaker 3: thing as point three repeating, and so they know that, 388 00:17:09,560 --> 00:17:11,800 Speaker 3: Like when your calculator shows zero point three two two 389 00:17:11,800 --> 00:17:13,639 Speaker 3: two two two two two three, they're like, oh, one third, 390 00:17:14,000 --> 00:17:16,399 Speaker 3: And that's a simple example, but you could imagine other ones. 391 00:17:16,520 --> 00:17:18,800 Speaker 3: You could not send people to the moon if you 392 00:17:18,920 --> 00:17:22,400 Speaker 3: only use a calculator but you didn't have an understanding 393 00:17:22,400 --> 00:17:24,520 Speaker 3: of things like floating point errors, or if you didn't 394 00:17:24,600 --> 00:17:27,720 Speaker 3: understand things of whereas the calculator have a limitation. Similar thing. 395 00:17:28,440 --> 00:17:31,000 Speaker 3: Internet comes out, Google comes out. People say, oh, you 396 00:17:31,040 --> 00:17:33,639 Speaker 3: can web search everything. Why do you need to have 397 00:17:33,680 --> 00:17:37,720 Speaker 3: any knowledge anymore? It turns out that people with knowledge 398 00:17:37,760 --> 00:17:40,359 Speaker 3: are going to have much better information. For example, you 399 00:17:40,400 --> 00:17:42,120 Speaker 3: might have forgotten that it was the Battle of Hastings, 400 00:17:42,440 --> 00:17:46,200 Speaker 3: but you're you'll be like, hey, you know google, William 401 00:17:46,280 --> 00:17:47,600 Speaker 3: the Conqueror, what was that battle again? 402 00:17:48,720 --> 00:17:51,199 Speaker 2: Or you know, I remember that the card English. 403 00:17:51,280 --> 00:17:53,240 Speaker 3: You know, English has a lot of French words in 404 00:17:53,280 --> 00:17:56,199 Speaker 3: it was it was that related to the Battle of 405 00:17:56,200 --> 00:18:00,040 Speaker 3: Hastings and and and the Norman conquest. And if you 406 00:18:00,160 --> 00:18:02,439 Speaker 3: at least have some of that, you wouldn't even begin 407 00:18:02,600 --> 00:18:04,720 Speaker 3: to ask the right questions. Or let's say that you 408 00:18:04,800 --> 00:18:06,680 Speaker 3: just have a homework assignment and you copy and paste 409 00:18:06,680 --> 00:18:09,280 Speaker 3: it into Google this is twenty years ago. You got 410 00:18:09,520 --> 00:18:12,960 Speaker 3: you got five search results, or YouTube videos start showing up. 411 00:18:13,720 --> 00:18:16,760 Speaker 3: Unless you have some context and some critical thinking of 412 00:18:16,840 --> 00:18:21,080 Speaker 3: judging what's solid information or not, you could just be 413 00:18:21,080 --> 00:18:22,639 Speaker 3: completely misguided. Like if you didn't know it was the 414 00:18:22,640 --> 00:18:24,960 Speaker 3: Battle of Hastings and you click on a video it's 415 00:18:25,000 --> 00:18:26,679 Speaker 3: like we're going to talk about William the Conqueror and 416 00:18:26,720 --> 00:18:29,720 Speaker 3: the Battle of Pudding and you're like, oh, it's the 417 00:18:29,720 --> 00:18:32,359 Speaker 3: Battle of Pudding. And it might have been a joke video, 418 00:18:32,440 --> 00:18:34,119 Speaker 3: but you didn't even watch enough of it to like 419 00:18:34,480 --> 00:18:37,520 Speaker 3: that's going to be a problem. And I see pre AI, 420 00:18:37,720 --> 00:18:40,000 Speaker 3: those with the most content knowledge are the people in 421 00:18:40,000 --> 00:18:42,960 Speaker 3: the position to use the Internet for their knowledge work, 422 00:18:43,160 --> 00:18:45,480 Speaker 3: and they're the ones that I see the difference between 423 00:18:46,080 --> 00:18:48,520 Speaker 3: when I web search versus when a younger child does, 424 00:18:48,600 --> 00:18:51,800 Speaker 3: or even some people you know in my family who 425 00:18:51,880 --> 00:18:55,000 Speaker 3: might not have the full context of what they're searching for. 426 00:18:55,320 --> 00:18:57,639 Speaker 3: They're doing very superficial searches. They don't know the right 427 00:18:57,680 --> 00:18:59,960 Speaker 3: questions to ask. Same thing with AI. AI is like 428 00:19:00,480 --> 00:19:04,159 Speaker 3: can be an intelligent partner with you. Imagine trying to 429 00:19:04,200 --> 00:19:07,720 Speaker 3: have a philosophical conversation with Aristotle without having read any 430 00:19:07,720 --> 00:19:10,560 Speaker 3: of Aristotle, or without having read any philosophy, or at 431 00:19:10,640 --> 00:19:12,840 Speaker 3: least without having talked about it. Or imagine you know, 432 00:19:12,880 --> 00:19:16,240 Speaker 3: being able to meet Isaac Newton, but you don't. You 433 00:19:16,560 --> 00:19:20,080 Speaker 3: haven't gotten the basics of physics or calculus down. You 434 00:19:20,119 --> 00:19:21,800 Speaker 3: can't just tell them like, what's going to happen to 435 00:19:21,840 --> 00:19:24,800 Speaker 3: this rock? If you should say, hey, does Newton's third 436 00:19:24,880 --> 00:19:26,760 Speaker 3: law apply here? Because this seems like an edge case, 437 00:19:26,800 --> 00:19:27,879 Speaker 3: and then you're going to be able to go much 438 00:19:27,920 --> 00:19:30,480 Speaker 3: much further. So I think content knowledge is actually more 439 00:19:30,520 --> 00:19:33,520 Speaker 3: important now because if you want to leverage AI, if 440 00:19:33,520 --> 00:19:35,199 Speaker 3: you want it to amplify your intent, you have to 441 00:19:35,240 --> 00:19:36,000 Speaker 3: hang with the AI. 442 00:19:36,280 --> 00:19:38,600 Speaker 2: And there's still going to be work, human centered work. 443 00:19:38,640 --> 00:19:41,040 Speaker 3: There's going to be taking care of seniors, things that 444 00:19:41,040 --> 00:19:44,840 Speaker 3: are very more manual or more human driven hospitality. But 445 00:19:44,880 --> 00:19:47,159 Speaker 3: I think the knowledge economy where AI is going to 446 00:19:47,160 --> 00:19:48,760 Speaker 3: become a big part of it, you have to be 447 00:19:48,800 --> 00:19:50,720 Speaker 3: able to hang with it, and content will matter more. 448 00:19:51,000 --> 00:19:54,280 Speaker 1: I agree with you, although I do wonder about memorization 449 00:19:54,600 --> 00:19:57,840 Speaker 1: versus critical thinking. So how do you think about teaching 450 00:19:57,920 --> 00:19:59,120 Speaker 1: children critical thinking? 451 00:19:59,640 --> 00:20:02,560 Speaker 2: At this point here, I'll also sound a little traditionalist. 452 00:20:02,600 --> 00:20:03,240 Speaker 2: I think it's both. 453 00:20:03,680 --> 00:20:07,520 Speaker 3: I think it's become very I definitely disagree, like I 454 00:20:07,520 --> 00:20:11,200 Speaker 3: don't like pure like Victorian era memorize everything like that's 455 00:20:11,240 --> 00:20:15,240 Speaker 3: not good. But I do think in like Western school systems, 456 00:20:15,240 --> 00:20:19,239 Speaker 3: it's become fashionable to say any type of memorization is 457 00:20:19,280 --> 00:20:22,040 Speaker 3: not good, and we're only going to focus on critical 458 00:20:22,040 --> 00:20:24,960 Speaker 3: thinking and skill development. And the reality is is that 459 00:20:25,720 --> 00:20:27,520 Speaker 3: I'll speak for myself, A lot of the connections that 460 00:20:27,520 --> 00:20:31,040 Speaker 3: have formed in my mind are because I knew something 461 00:20:31,880 --> 00:20:35,080 Speaker 3: and then my brain at some point, Like you know, 462 00:20:35,200 --> 00:20:37,920 Speaker 3: I every American kid learns about the Louisiana purchase in 463 00:20:37,960 --> 00:20:40,640 Speaker 3: eighteen oh three, and then later most American kids don't 464 00:20:40,680 --> 00:20:42,520 Speaker 3: learn about the Napoleonic Wars, even though that was a 465 00:20:42,600 --> 00:20:43,480 Speaker 3: much more important thing. 466 00:20:43,960 --> 00:20:45,760 Speaker 2: But if you do learn about the Napoleonic Wars, and 467 00:20:45,760 --> 00:20:46,600 Speaker 2: they're normally. 468 00:20:46,280 --> 00:20:50,240 Speaker 3: Not connected for anyone right, But at some point you're like, oh, wait, 469 00:20:50,720 --> 00:20:53,200 Speaker 3: in that other class, I learned like he's sitting here 470 00:20:53,240 --> 00:20:56,320 Speaker 3: fighting all of these other European powers, and eighteen oh 471 00:20:56,320 --> 00:21:00,280 Speaker 3: three he actually his navy was defeated. I wonder whether 472 00:21:00,280 --> 00:21:03,399 Speaker 3: that has anything to do with the Louisiana purchase. And 473 00:21:03,480 --> 00:21:05,119 Speaker 3: even though they never teach us you in history class, 474 00:21:05,119 --> 00:21:07,760 Speaker 3: has everything to do with the Louisiana purchase. Napoleon sold 475 00:21:07,800 --> 00:21:09,560 Speaker 3: it because there's no way he could defend it. It turns 476 00:21:09,600 --> 00:21:12,800 Speaker 3: out Thomas Jefferson could have just taken Louisiana and Napoleon 477 00:21:12,840 --> 00:21:15,320 Speaker 3: had no navy to send out because the British had 478 00:21:15,359 --> 00:21:19,080 Speaker 3: just destroyed it. And that's critical thinking. But in order 479 00:21:19,119 --> 00:21:24,040 Speaker 3: to create those connections, you had to put the Napoleonic 480 00:21:24,080 --> 00:21:27,679 Speaker 3: Wars in time and space with the Louisiana purchase, and 481 00:21:27,720 --> 00:21:29,280 Speaker 3: you have to know where these things are. You have 482 00:21:29,320 --> 00:21:32,199 Speaker 3: to know where Napoleon is, where England is, where the 483 00:21:32,280 --> 00:21:35,359 Speaker 3: navy was defeated, where Louisiana is. You're like, oh, critical thinking, 484 00:21:35,720 --> 00:21:38,359 Speaker 3: I'd be awfully hard to defend Louisiana if I'm sitting 485 00:21:38,359 --> 00:21:40,440 Speaker 3: here battling all these Europeans and I don't have no 486 00:21:40,520 --> 00:21:41,160 Speaker 3: navy anymore. 487 00:21:55,320 --> 00:21:56,120 Speaker 2: Well, let me see two. 488 00:21:56,040 --> 00:21:58,440 Speaker 1: Things from the brain point of view, which is four 489 00:21:58,560 --> 00:22:03,199 Speaker 1: plasticity to happen. For knowledge to stick, it has to 490 00:22:03,240 --> 00:22:06,199 Speaker 1: be relevant in some way. And so one of the 491 00:22:06,200 --> 00:22:08,280 Speaker 1: ways that knowledge becomes relevant for us is when we 492 00:22:08,320 --> 00:22:10,359 Speaker 1: are curious about something. So if you think, hey, wait, 493 00:22:10,520 --> 00:22:12,320 Speaker 1: did the employee ack wars have anything to do with it? 494 00:22:12,400 --> 00:22:14,040 Speaker 1: Then you're curious and you look it up and then 495 00:22:14,080 --> 00:22:15,120 Speaker 1: it sticks the rest of your life. 496 00:22:15,240 --> 00:22:15,520 Speaker 2: Great. 497 00:22:15,960 --> 00:22:20,000 Speaker 1: But how do we, as educators and using AI, how 498 00:22:20,000 --> 00:22:22,960 Speaker 1: do we make things relevant to kids? And how do 499 00:22:23,119 --> 00:22:25,679 Speaker 1: two AI tutors play part in that? Yeah, That's what 500 00:22:25,680 --> 00:22:26,440 Speaker 1: I'm excited about. 501 00:22:26,600 --> 00:22:29,399 Speaker 3: And I'll keep giving display because I don't want to 502 00:22:29,440 --> 00:22:31,240 Speaker 3: make it sound like AI to mar Is is going to 503 00:22:31,280 --> 00:22:34,160 Speaker 3: make everything perfect. But we're on a path and we've 504 00:22:34,200 --> 00:22:37,119 Speaker 3: already seen once again that fifteen twenty percent of students 505 00:22:37,119 --> 00:22:40,240 Speaker 3: who are let's call it proactive and showing more of 506 00:22:40,240 --> 00:22:42,000 Speaker 3: that curiosity. I've seen it in front of my eyes 507 00:22:42,000 --> 00:22:44,679 Speaker 3: where they're learning about the amendments of the Constitution. But 508 00:22:44,720 --> 00:22:46,439 Speaker 3: now with generative AI, they could say, hey, why is 509 00:22:46,440 --> 00:22:48,240 Speaker 3: this relevant now? And then the IA says, well, actually, 510 00:22:49,119 --> 00:22:51,920 Speaker 3: back just recently in twenty twenty there was this Supreme 511 00:22:51,960 --> 00:22:54,679 Speaker 3: Court case and actually that makes a big difference on 512 00:22:54,840 --> 00:22:57,359 Speaker 3: like executive power, the Chevron doctrine, and. 513 00:22:57,359 --> 00:22:59,280 Speaker 2: Are like, oh wow, this is a big deal. I 514 00:22:59,359 --> 00:22:59,919 Speaker 2: never realized. 515 00:23:00,440 --> 00:23:02,600 Speaker 3: So there's apps or like hey, I'm studying entroped, why 516 00:23:02,600 --> 00:23:04,320 Speaker 3: does this matter? And it's like, no, it actually matters 517 00:23:04,320 --> 00:23:06,480 Speaker 3: a ton because if you go into this or there's 518 00:23:06,560 --> 00:23:09,280 Speaker 3: just a study recently, or if you're trying to and 519 00:23:09,560 --> 00:23:11,560 Speaker 3: so I think that's one of the really powerful things 520 00:23:11,600 --> 00:23:14,320 Speaker 3: that it. You know, the classic question for every kid, 521 00:23:14,440 --> 00:23:16,560 Speaker 3: when am I going to need to use this? I 522 00:23:16,600 --> 00:23:18,480 Speaker 3: consider myself better than most people and being able to 523 00:23:18,520 --> 00:23:21,679 Speaker 3: answer that question, but even I'll fumble sometimes. But the 524 00:23:21,680 --> 00:23:23,920 Speaker 3: AI is very good at that, and it's very good 525 00:23:23,920 --> 00:23:26,959 Speaker 3: at like on conmego, if a student literally asked why 526 00:23:27,000 --> 00:23:28,199 Speaker 3: do I need to learn this? The first thing that 527 00:23:28,200 --> 00:23:29,840 Speaker 3: the AI typically does is is, well, what do you 528 00:23:29,840 --> 00:23:31,800 Speaker 3: care about? And the students I want to be a 529 00:23:31,840 --> 00:23:33,639 Speaker 3: professional athlete, I want to do this, and it'll actually 530 00:23:33,640 --> 00:23:35,520 Speaker 3: the I will remember that there's ways to reset it 531 00:23:35,560 --> 00:23:37,679 Speaker 3: and clear it and so it's not creepy. But like 532 00:23:38,200 --> 00:23:40,840 Speaker 3: then in the future, it can proactively say, well, you know, 533 00:23:40,880 --> 00:23:42,240 Speaker 3: since you said you want to be an athlete. 534 00:23:42,640 --> 00:23:45,040 Speaker 2: This molecule, you know, ATP is crucial. 535 00:23:45,520 --> 00:23:47,280 Speaker 3: You know, you want to maximize your ATP if you're 536 00:23:47,280 --> 00:23:49,480 Speaker 3: about to you know, run a marathon or something. 537 00:23:49,800 --> 00:23:51,480 Speaker 2: So that's the hope. 538 00:23:51,640 --> 00:23:54,560 Speaker 3: And as AI gets better and better, as it gets multimodal, 539 00:23:54,560 --> 00:23:58,000 Speaker 3: as it gets you know, we can completely imagine ten 540 00:23:58,080 --> 00:24:00,200 Speaker 3: years in the future, for sure you're going to You'll 541 00:24:00,240 --> 00:24:03,439 Speaker 3: put on augmented reality glasses or virtual reality glasses and 542 00:24:04,080 --> 00:24:06,679 Speaker 3: travel to ancient Rome and try to stop, you know, 543 00:24:06,720 --> 00:24:09,919 Speaker 3: the assassination of Caesar, or travel into the human brain 544 00:24:10,320 --> 00:24:12,600 Speaker 3: and try to, you know, diagnose whether this is a 545 00:24:12,960 --> 00:24:15,760 Speaker 3: you know, an Alzheimer's case, or you know, are there 546 00:24:15,800 --> 00:24:17,159 Speaker 3: ways that you might be able to address it. So 547 00:24:17,520 --> 00:24:20,439 Speaker 3: I think that relevance question AI will help. It's not 548 00:24:20,680 --> 00:24:23,800 Speaker 3: solved it now. I'll say one other thing though, you know, 549 00:24:23,840 --> 00:24:26,040 Speaker 3: I in my first book, I gave the example of 550 00:24:26,400 --> 00:24:28,879 Speaker 3: you know, people always talk about relevance, but I was 551 00:24:28,920 --> 00:24:31,000 Speaker 3: this I don't know. I used to think I was 552 00:24:31,000 --> 00:24:33,000 Speaker 3: five nine, but I now realize probably closer to five 553 00:24:33,000 --> 00:24:33,639 Speaker 3: to seven. 554 00:24:33,560 --> 00:24:34,000 Speaker 2: Five eight. 555 00:24:34,200 --> 00:24:37,919 Speaker 3: But I was this that tall kid in high school. 556 00:24:38,160 --> 00:24:40,359 Speaker 3: But when I was in Pe, I never said, why 557 00:24:40,359 --> 00:24:42,160 Speaker 3: am I going to have to take an orange ball 558 00:24:42,200 --> 00:24:44,159 Speaker 3: and have to put it into a ten foot hoop, 559 00:24:44,240 --> 00:24:46,320 Speaker 3: and this has no relev It had no relevance to 560 00:24:46,640 --> 00:24:49,440 Speaker 3: my but I was engaged because my friends were there, 561 00:24:49,520 --> 00:24:53,080 Speaker 3: I was having fun. I didn't feel disengaged. I do 562 00:24:53,119 --> 00:24:56,520 Speaker 3: think that sometimes the relevance question is missing the mark 563 00:24:56,560 --> 00:24:59,040 Speaker 3: a bit like math to some degree, the real beauty 564 00:24:59,080 --> 00:25:01,040 Speaker 3: of math isn't in its ow implication. It's in its 565 00:25:01,040 --> 00:25:04,159 Speaker 3: inherent beauty in it. It's the thought process of it. 566 00:25:04,400 --> 00:25:05,680 Speaker 3: And I think the reason why a lot of kids 567 00:25:05,680 --> 00:25:07,800 Speaker 3: start saying, oh, when am I going to learn this 568 00:25:08,400 --> 00:25:10,600 Speaker 3: is they have so many gaps that they've disengaged, so 569 00:25:10,600 --> 00:25:12,880 Speaker 3: they're just protecting their self esteem if you can fill 570 00:25:12,880 --> 00:25:14,480 Speaker 3: in those gaps. I saw this with my cousins. Some 571 00:25:14,520 --> 00:25:16,760 Speaker 3: of my cousins were saying, I'm not a math person. 572 00:25:16,840 --> 00:25:17,440 Speaker 3: I hate math. 573 00:25:17,560 --> 00:25:20,280 Speaker 2: Is that. But once they started getting that personal attention 574 00:25:20,320 --> 00:25:22,640 Speaker 2: and I'm like, Hey, isn't this cool? Isn't this cool? 575 00:25:22,680 --> 00:25:24,600 Speaker 3: You can do it this way or this way? Isn't 576 00:25:24,600 --> 00:25:27,280 Speaker 3: this amazing that this can be described or predicted? They 577 00:25:27,280 --> 00:25:29,359 Speaker 3: started becoming math people. And not that they thought that 578 00:25:29,400 --> 00:25:30,399 Speaker 3: they were going to use it in their life, but 579 00:25:30,400 --> 00:25:31,760 Speaker 3: they just started seeing the beauty in it and they 580 00:25:31,800 --> 00:25:33,040 Speaker 3: start feeling confident in it. 581 00:25:33,240 --> 00:25:34,199 Speaker 2: That's great. You know. 582 00:25:34,280 --> 00:25:37,679 Speaker 1: Another thing is we as humans are so sensitive to 583 00:25:37,920 --> 00:25:41,520 Speaker 1: story as opposed to let's say, a bulleted list of points. 584 00:25:42,000 --> 00:25:44,080 Speaker 1: And so this is one of the things I'm enthusiastic about 585 00:25:44,080 --> 00:25:47,320 Speaker 1: with AI is can it take complicated things, whether it's 586 00:25:47,400 --> 00:25:49,560 Speaker 1: history or math or whatever and tell you a story 587 00:25:49,600 --> 00:25:50,080 Speaker 1: about it. 588 00:25:50,440 --> 00:25:53,120 Speaker 3: Absolutely, I can create a story. It could create a poem, 589 00:25:53,280 --> 00:25:57,160 Speaker 3: a song. It can. I think it can create simulations 590 00:25:57,440 --> 00:26:00,720 Speaker 3: for you where yes, it becomes like the ultimate choose 591 00:26:00,720 --> 00:26:01,960 Speaker 3: your own eventure. I mean, right now, it could be 592 00:26:01,960 --> 00:26:04,760 Speaker 3: text based, it's soon going to be a video and 593 00:26:05,359 --> 00:26:08,200 Speaker 3: visual based. But as we said in the not too 594 00:26:08,240 --> 00:26:11,280 Speaker 3: far off future, let's call it five ten years, it's 595 00:26:11,280 --> 00:26:12,679 Speaker 3: going to be able to create a lucid dream for 596 00:26:12,720 --> 00:26:15,160 Speaker 3: you that puts you in the middle of it all. 597 00:26:15,640 --> 00:26:15,960 Speaker 2: Yeah. 598 00:26:16,119 --> 00:26:17,560 Speaker 1: So I want to come back to this point about 599 00:26:17,600 --> 00:26:20,639 Speaker 1: what this means for teachers. Tell us your thoughts on that. 600 00:26:21,119 --> 00:26:24,399 Speaker 3: I've been saying this for years because even when I 601 00:26:24,440 --> 00:26:27,399 Speaker 3: started making videos and software, some people said, oh, on 602 00:26:27,480 --> 00:26:31,879 Speaker 3: demand video does this replace teachers? And my response is 603 00:26:31,960 --> 00:26:34,960 Speaker 3: I don't think it does. If all a teacher is 604 00:26:35,040 --> 00:26:40,600 Speaker 3: doing is going up, giving a lecture and leaving. Then yes, 605 00:26:40,680 --> 00:26:43,960 Speaker 3: maybe on demand video could replace that. And you see 606 00:26:44,000 --> 00:26:45,760 Speaker 3: that happening in a lot of colleges. They've you know, 607 00:26:45,800 --> 00:26:47,119 Speaker 3: no one goes in med schools, no one goes to 608 00:26:47,119 --> 00:26:49,520 Speaker 3: the lecture anymore, because that's what the lectures were. People 609 00:26:49,560 --> 00:26:52,439 Speaker 3: are watching the recordings of the lectures on double speed, 610 00:26:52,720 --> 00:26:54,240 Speaker 3: and now they can ask questions to an AI. 611 00:26:55,359 --> 00:26:55,920 Speaker 2: So that's that. 612 00:26:56,040 --> 00:26:58,199 Speaker 3: But and I've been saying this for years, the ideal 613 00:26:58,240 --> 00:27:01,119 Speaker 3: classroom has always been, and this go back to Socrates 614 00:27:02,200 --> 00:27:05,760 Speaker 3: and Plato and Aristotle, is if human beings are getting 615 00:27:05,760 --> 00:27:08,800 Speaker 3: together in a room, make them interact with each other, 616 00:27:09,320 --> 00:27:12,520 Speaker 3: have a Socratic dialogue, have a simulation. Have students go 617 00:27:12,560 --> 00:27:15,080 Speaker 3: into groups and tackle the problem a little bit. Yes, 618 00:27:15,119 --> 00:27:16,680 Speaker 3: you could give a little bit of a mini lecture, 619 00:27:17,080 --> 00:27:20,840 Speaker 3: but then make it interactive or maybe out of the 620 00:27:20,840 --> 00:27:23,440 Speaker 3: thirty students, ten of them really need help with this concept, 621 00:27:23,440 --> 00:27:25,040 Speaker 3: while the other twenty can actually learn at their own 622 00:27:25,040 --> 00:27:27,560 Speaker 3: pace or keep working on their project, or work on connuiccuting, whatever. 623 00:27:27,800 --> 00:27:30,199 Speaker 3: So do the focus intervention with these ten, and then 624 00:27:30,240 --> 00:27:32,560 Speaker 3: you don't have to bore or lose the other twenty. 625 00:27:32,600 --> 00:27:34,600 Speaker 3: And then do another focus intervention with the other ten 626 00:27:35,040 --> 00:27:38,080 Speaker 3: and that type of work. I think most people go 627 00:27:38,119 --> 00:27:40,080 Speaker 3: into the teaching profession because they actually want to be 628 00:27:40,119 --> 00:27:44,200 Speaker 3: that teacher that unlocks little sal or, little David, makes 629 00:27:44,200 --> 00:27:47,399 Speaker 3: them believe in themselves. You know, we all can list 630 00:27:47,880 --> 00:27:50,600 Speaker 3: five to ten teachers at every stage of our life. 631 00:27:50,640 --> 00:27:52,480 Speaker 3: It's like that was the teacher that made me believe 632 00:27:52,480 --> 00:27:55,200 Speaker 3: in myself or made me see wonder in the world, 633 00:27:55,480 --> 00:27:57,760 Speaker 3: and they change my trajectory. And every teacher wants to 634 00:27:57,800 --> 00:28:00,920 Speaker 3: be that teacher. And I think whether it's video, whether 635 00:28:00,960 --> 00:28:04,000 Speaker 3: it's software, and it's AI, if it's used well and 636 00:28:04,040 --> 00:28:05,840 Speaker 3: the AI is acting as assistant but not trying to 637 00:28:05,920 --> 00:28:07,919 Speaker 3: replace the teacher, which I don't think it should. I 638 00:28:07,920 --> 00:28:09,439 Speaker 3: think it could actually free up the teacher to do 639 00:28:09,480 --> 00:28:11,440 Speaker 3: a lot of that more human element to work. 640 00:28:11,840 --> 00:28:13,119 Speaker 1: And I know one of the things that you've been 641 00:28:13,160 --> 00:28:17,480 Speaker 1: looking at is how AI could improve debating, as in, 642 00:28:17,560 --> 00:28:18,720 Speaker 1: I'm going to take a point. 643 00:28:18,480 --> 00:28:19,840 Speaker 2: And the AI is going to give you the opposite 644 00:28:19,880 --> 00:28:20,840 Speaker 2: point and tell us about that. 645 00:28:21,080 --> 00:28:23,520 Speaker 3: Yeah, and it's you know, there's actually two projects on 646 00:28:23,560 --> 00:28:26,840 Speaker 3: the debating let's call it debate, and one is very 647 00:28:26,880 --> 00:28:28,840 Speaker 3: AI and one is very non AI. But I think 648 00:28:28,880 --> 00:28:32,000 Speaker 3: they might actually converge. So on Conmigo, we actually have 649 00:28:32,080 --> 00:28:34,199 Speaker 3: an activity where you can get into a debate with 650 00:28:34,200 --> 00:28:37,080 Speaker 3: an AI on anything, you know, should the government cancel 651 00:28:37,160 --> 00:28:40,560 Speaker 3: student debt, should the US maintain military bases around the world, 652 00:28:40,640 --> 00:28:43,160 Speaker 3: or should it you know, kind of become more isolationist 653 00:28:43,600 --> 00:28:46,360 Speaker 3: and you can pick either side. The AI will have 654 00:28:46,400 --> 00:28:48,480 Speaker 3: a polite debate and then actually give you feedback and 655 00:28:48,480 --> 00:28:50,280 Speaker 3: then you can switch side. So that's where we've got 656 00:28:50,280 --> 00:28:53,040 Speaker 3: a lot of positive feedback. Students enjoy that it's a 657 00:28:53,080 --> 00:28:55,840 Speaker 3: safe environment to test ideas without being embarrassed. 658 00:28:56,240 --> 00:28:57,360 Speaker 2: Now we have another project. 659 00:28:57,400 --> 00:28:59,960 Speaker 3: There's a sister nonprofit to Kind Academy that we still 660 00:29:00,120 --> 00:29:03,200 Speaker 3: during the pandemic called Schoolhouse dot World. Schoolhouse dot World 661 00:29:03,280 --> 00:29:06,719 Speaker 3: is all about free live tutoring from other human beings. 662 00:29:07,520 --> 00:29:09,960 Speaker 3: The way that we're able to give people free live 663 00:29:10,040 --> 00:29:13,120 Speaker 3: over zoom tutoring from other human beings is through volunteers, 664 00:29:13,360 --> 00:29:15,960 Speaker 3: and anyone listening can go be a volunteer, and there's 665 00:29:15,960 --> 00:29:19,200 Speaker 3: a whole mechanism to certify yourself or you can get tutoring. 666 00:29:19,640 --> 00:29:21,520 Speaker 3: But one of the things we realize a lot of 667 00:29:21,600 --> 00:29:27,160 Speaker 3: universities and some of the top universities you, Chicago, MIT, Brown, Yale, Caltech. 668 00:29:27,560 --> 00:29:30,640 Speaker 3: There's thirty others said, Hey, if someone is tutoring on 669 00:29:30,680 --> 00:29:33,120 Speaker 3: Schoolhouse dot World, so they've certified that they say no 670 00:29:33,200 --> 00:29:35,600 Speaker 3: physics or they know biology, and they're tutoring other people, 671 00:29:35,960 --> 00:29:38,560 Speaker 3: we want to know about that for college admissions. And 672 00:29:38,600 --> 00:29:40,520 Speaker 3: so they have been using it for college admissions. But 673 00:29:40,520 --> 00:29:42,920 Speaker 3: on top of that, when we went last year, they're like, hey, 674 00:29:42,920 --> 00:29:44,520 Speaker 3: is there any way that we could leverage this platform 675 00:29:44,560 --> 00:29:49,479 Speaker 3: to foster like civil dialogue between people, civil debate to 676 00:29:49,520 --> 00:29:51,640 Speaker 3: your point, And we're like, okay, let's try something a 677 00:29:51,680 --> 00:29:54,920 Speaker 3: little bit wild. Let's have real people fill out a 678 00:29:54,960 --> 00:29:59,000 Speaker 3: survey on all the hottest button issues Israel, Palestine, abortion, 679 00:29:59,400 --> 00:30:04,080 Speaker 3: affirmative action, free speech, gun control, gun control, go down 680 00:30:04,080 --> 00:30:06,040 Speaker 3: the list, the stuff that people are afraid to talk 681 00:30:06,040 --> 00:30:07,640 Speaker 3: about because they don't want to lose friendships. 682 00:30:08,200 --> 00:30:09,040 Speaker 2: Students fill it out. 683 00:30:09,200 --> 00:30:11,920 Speaker 3: You get paired, either one on one, sometimes two on 684 00:30:11,960 --> 00:30:16,000 Speaker 3: two with someone with a diametrically opposite point of view. 685 00:30:16,400 --> 00:30:19,760 Speaker 3: Now what happens on this dialogue project is you're not 686 00:30:19,760 --> 00:30:22,360 Speaker 3: supposed to it's not a debate, but you're supposed to 687 00:30:22,400 --> 00:30:24,520 Speaker 3: listen to each other. And at the end of it, 688 00:30:25,040 --> 00:30:27,959 Speaker 3: you have to with the form represent the other person's 689 00:30:28,000 --> 00:30:30,240 Speaker 3: point of view. You also have to reflect on whether 690 00:30:30,280 --> 00:30:32,840 Speaker 3: you think they heard you and represent and then any 691 00:30:32,880 --> 00:30:35,560 Speaker 3: other reflections on the interaction that you had. And what 692 00:30:35,680 --> 00:30:38,280 Speaker 3: happens then is you get a rating on your ability 693 00:30:38,280 --> 00:30:42,480 Speaker 3: to engage in hard conversations. And these colleges are very 694 00:30:42,520 --> 00:30:44,480 Speaker 3: interested in this as they don't carry your point of 695 00:30:44,520 --> 00:30:46,480 Speaker 3: view on these issues. They care that they like that 696 00:30:46,520 --> 00:30:48,600 Speaker 3: you have a point of view, but that you were 697 00:30:48,640 --> 00:30:50,640 Speaker 3: able to engage in a respectful and thoughtful and you're 698 00:30:50,680 --> 00:30:52,760 Speaker 3: able to steal man the other person's argument, you're able 699 00:30:52,840 --> 00:30:55,840 Speaker 3: to represent it. Now, those are both ways to have 700 00:30:55,960 --> 00:31:00,000 Speaker 3: dialogue in a world where it's getting increasingly polarized, increasingly 701 00:31:00,120 --> 00:31:03,240 Speaker 3: hard to have these conversations and feel safe. Over time, 702 00:31:03,280 --> 00:31:05,440 Speaker 3: these might converge a little bit. We are imagining ways 703 00:31:05,440 --> 00:31:08,400 Speaker 3: that AI. We are already using AI on schoolhouse to 704 00:31:08,440 --> 00:31:11,080 Speaker 3: give the tutor's feedback after the fact to saying, hey, 705 00:31:11,160 --> 00:31:13,920 Speaker 3: you know, you probably didn't ask enough questions, or you 706 00:31:13,960 --> 00:31:16,840 Speaker 3: ignored this one student, or they were actually asking this. 707 00:31:17,320 --> 00:31:19,360 Speaker 3: Over time is going to be real time. We also 708 00:31:19,400 --> 00:31:22,040 Speaker 3: imagine this dialogue project the AI might be able to facilitate. 709 00:31:22,080 --> 00:31:24,240 Speaker 3: We actually haven't had any where people started yelling at 710 00:31:24,280 --> 00:31:26,760 Speaker 3: each other or got on out. But maybe it can 711 00:31:26,800 --> 00:31:29,959 Speaker 3: make sure that the conversation. Oh oh, David, you know 712 00:31:30,200 --> 00:31:31,760 Speaker 3: I can tell your feeling a little sensitive about what 713 00:31:31,840 --> 00:31:33,280 Speaker 3: Sala said. Sal, is there another way that you might 714 00:31:33,320 --> 00:31:35,360 Speaker 3: have said that that you know might be a little 715 00:31:35,440 --> 00:31:35,959 Speaker 3: less triggering? 716 00:31:36,080 --> 00:31:38,440 Speaker 2: Or David, can you tell sal a way. 717 00:31:38,240 --> 00:31:40,360 Speaker 3: That he might have said that that you would have 718 00:31:40,360 --> 00:31:43,120 Speaker 3: appreciated things like that? I could imagine AI facilitating that 719 00:31:43,120 --> 00:31:43,640 Speaker 3: type of thing. 720 00:31:43,800 --> 00:31:45,959 Speaker 1: Oh great, And what do you think is the future 721 00:31:46,000 --> 00:31:49,800 Speaker 1: of AI in terms of assessments instead of taking standardized tests? 722 00:31:50,240 --> 00:31:52,800 Speaker 1: What's the future in five years? 723 00:31:53,240 --> 00:31:53,840 Speaker 2: Like everything? 724 00:31:53,880 --> 00:31:55,640 Speaker 3: It's I mean, if if we're living in a science 725 00:31:55,640 --> 00:32:00,800 Speaker 3: fiction book and standardized tests I have historically defended because 726 00:32:00,800 --> 00:32:03,200 Speaker 3: I've said, look, to get good at anything, you have 727 00:32:03,240 --> 00:32:06,280 Speaker 3: to measure it, And would you prefer unstandardized measurement you 728 00:32:06,320 --> 00:32:08,560 Speaker 3: want to be able to benchmark. Now, the problem that 729 00:32:08,560 --> 00:32:10,719 Speaker 3: most people have with sanitized tests, which I also agree with, 730 00:32:10,840 --> 00:32:14,840 Speaker 3: is if you want something that's consistent and standardized and scalable, 731 00:32:15,280 --> 00:32:18,200 Speaker 3: it's got to be you know, fairly narrow in what 732 00:32:18,240 --> 00:32:21,040 Speaker 3: it can measure. So most standardized tests are multiple choice 733 00:32:21,120 --> 00:32:23,720 Speaker 3: or just numeric entry. They don't let you do a 734 00:32:23,760 --> 00:32:26,840 Speaker 3: lot of open ended things. And unfortunately, yes, they do 735 00:32:26,880 --> 00:32:31,080 Speaker 3: measure some things that are of value, but they can't 736 00:32:31,120 --> 00:32:33,120 Speaker 3: measure how well you could write a longer paper, or 737 00:32:33,160 --> 00:32:35,240 Speaker 3: how well you can design something, or how well you 738 00:32:35,280 --> 00:32:39,080 Speaker 3: can problem solve things. What's exciting about generative AI is yes, 739 00:32:39,120 --> 00:32:41,040 Speaker 3: I do think in the next in fact, we're working 740 00:32:41,040 --> 00:32:43,840 Speaker 3: on this kind of the future of assessment. We can 741 00:32:43,880 --> 00:32:47,520 Speaker 3: incrementally add layers on top of traditional standardized assessment. So 742 00:32:47,560 --> 00:32:50,000 Speaker 3: you can ask a student, Okay, you picked C. Can 743 00:32:50,000 --> 00:32:52,280 Speaker 3: you explain your reasoning? Why do you think you pick C? 744 00:32:53,080 --> 00:32:54,960 Speaker 3: Or even after you you actually got this one wrong, 745 00:32:55,000 --> 00:32:56,440 Speaker 3: do you think it was a careless mistake or do 746 00:32:56,440 --> 00:32:57,960 Speaker 3: you think you know there was an error? 747 00:32:57,960 --> 00:32:58,680 Speaker 2: What was your error? 748 00:32:58,920 --> 00:33:00,960 Speaker 3: And if you can give some of that subjective insights 749 00:33:00,960 --> 00:33:03,520 Speaker 3: to teachers, it becomes more actionable. Then they know what 750 00:33:03,560 --> 00:33:05,560 Speaker 3: to do with it. Reading comprehension instead of what was 751 00:33:05,600 --> 00:33:08,840 Speaker 3: the author's intent? You know ABCD, you could say, before 752 00:33:08,880 --> 00:33:11,080 Speaker 3: you see the choices in your own words, what do 753 00:33:11,120 --> 00:33:12,280 Speaker 3: you think the author's intent was? 754 00:33:12,640 --> 00:33:13,640 Speaker 2: Now look at the choices. 755 00:33:13,960 --> 00:33:16,760 Speaker 3: Now if you fast forward five years, ten years, I 756 00:33:16,760 --> 00:33:18,680 Speaker 3: think it's going to feel very much like a job 757 00:33:18,720 --> 00:33:22,280 Speaker 3: interview or like an oral thesis defense at a university. 758 00:33:22,800 --> 00:33:25,160 Speaker 3: And I know people's alarm bells start going off of 759 00:33:25,200 --> 00:33:27,200 Speaker 3: a bias and this and that and a it can 760 00:33:27,240 --> 00:33:28,840 Speaker 3: be creepy and data and all that, But I always 761 00:33:28,840 --> 00:33:34,000 Speaker 3: point out, like job interviews and these rich assessments like 762 00:33:34,120 --> 00:33:38,440 Speaker 3: thesis defense oral exams are rich, but they're incredibly biased, 763 00:33:38,600 --> 00:33:41,479 Speaker 3: and they're incredibly inconsistent. Right now, you know, even if 764 00:33:41,480 --> 00:33:44,400 Speaker 3: you're getting interviewed by the same person, depending on their mood, 765 00:33:44,800 --> 00:33:47,120 Speaker 3: depending on the day of the week, you could get 766 00:33:47,120 --> 00:33:49,760 Speaker 3: a completely different experience with that person. If we can 767 00:33:49,800 --> 00:33:51,840 Speaker 3: get AI, and if we can benchmark AI and test 768 00:33:51,840 --> 00:33:54,760 Speaker 3: AID and make sure, okay, two people equivalent, one is 769 00:33:54,800 --> 00:33:58,000 Speaker 3: African American, one is Asian, is giving the same signal 770 00:33:58,040 --> 00:34:02,560 Speaker 3: because other than race, these two simulated candidates are completely identical. 771 00:34:02,560 --> 00:34:04,280 Speaker 3: So you can benchmark an AI in ways that you 772 00:34:04,280 --> 00:34:06,640 Speaker 3: can't benchmark a human being. But if you can do that, 773 00:34:06,720 --> 00:34:08,320 Speaker 3: I think AI is going to You're going to be 774 00:34:08,320 --> 00:34:12,200 Speaker 3: able to have standardized assessments that feel a lot like 775 00:34:12,560 --> 00:34:16,279 Speaker 3: a job interview, that feel a lot like a simulation 776 00:34:16,520 --> 00:34:19,279 Speaker 3: or a game or a problem solving that I think 777 00:34:19,360 --> 00:34:21,400 Speaker 3: can really broaden the oppertunity. I mean even things like 778 00:34:21,440 --> 00:34:25,800 Speaker 3: sense of humor, or communication skills or body language. 779 00:34:25,880 --> 00:34:27,280 Speaker 2: I think the AI could give feedback. 780 00:34:27,320 --> 00:34:28,880 Speaker 3: Now, I'm a big believer you shouldn't just say, oh, 781 00:34:28,920 --> 00:34:31,479 Speaker 3: that person's body language was off, never give them a job. 782 00:34:31,840 --> 00:34:33,800 Speaker 3: It should be like, hey, you know what your body 783 00:34:33,880 --> 00:34:35,560 Speaker 3: language right now? You were a little bit you weren't 784 00:34:35,560 --> 00:34:37,839 Speaker 3: making eye contact, et cetera. I'd give you a three 785 00:34:37,880 --> 00:34:40,240 Speaker 3: out of five. Here's tips. Come back to me tomorrow 786 00:34:40,239 --> 00:34:42,200 Speaker 3: and take the assessment again. Mastery learning. 787 00:34:42,520 --> 00:34:45,320 Speaker 1: Oh great, so you've touched on a few things about 788 00:34:45,400 --> 00:34:47,359 Speaker 1: what the world might look like five ten years ago. 789 00:34:47,520 --> 00:34:50,200 Speaker 1: When I think of five years ago, I mean, the 790 00:34:50,239 --> 00:34:52,640 Speaker 1: only people who even knew these llms were working was 791 00:34:52,680 --> 00:34:54,439 Speaker 1: you know, some researchers that could will whatever. 792 00:34:54,480 --> 00:34:56,720 Speaker 2: But it hadn't burst onto the world yet. 793 00:34:57,239 --> 00:35:00,000 Speaker 1: And so five years hence, what else do you think 794 00:35:00,040 --> 00:35:02,040 Speaker 1: it's going to be different about, for example, what you're 795 00:35:02,040 --> 00:35:04,000 Speaker 1: doing with con migo or whatever is next. 796 00:35:04,040 --> 00:35:06,120 Speaker 2: What does it look like for students? It's hard. 797 00:35:06,360 --> 00:35:08,359 Speaker 3: Yeah, and you said five years ago. Even five years ago, 798 00:35:08,360 --> 00:35:11,799 Speaker 3: the AI researchers themselves didn't even appreciate I remember when 799 00:35:12,480 --> 00:35:14,840 Speaker 3: Greg and Sam gave us access to GPT four in 800 00:35:14,880 --> 00:35:17,279 Speaker 3: the summer of twenty twenty two, and my first question 801 00:35:17,320 --> 00:35:18,520 Speaker 3: before I even tried it out, I was like, does 802 00:35:18,560 --> 00:35:20,520 Speaker 3: this work in other languages? And they said, oh, we 803 00:35:20,520 --> 00:35:24,239 Speaker 3: don't think so. And then I barely speak Bengali, which 804 00:35:24,280 --> 00:35:26,920 Speaker 3: is where my family. My family comes from Bengal, and 805 00:35:27,000 --> 00:35:30,000 Speaker 3: I said some it wrote back to me to Bengali, 806 00:35:30,040 --> 00:35:32,280 Speaker 3: and I took a screenshot I sent to say They're like, oh, 807 00:35:32,400 --> 00:35:34,320 Speaker 3: oh yeah, we just figured that out when you asked, 808 00:35:34,600 --> 00:35:36,640 Speaker 3: so they didn't even know. And then one of my 809 00:35:36,960 --> 00:35:40,759 Speaker 3: friend's parents as a researcher at another AI firm, and 810 00:35:41,200 --> 00:35:43,759 Speaker 3: he's like, you know what, we didn't realize that you 811 00:35:43,800 --> 00:35:45,239 Speaker 3: guys did Kin Academy did. 812 00:35:45,280 --> 00:35:47,080 Speaker 2: I'm like, oh wow, what's our role here? 813 00:35:47,400 --> 00:35:49,640 Speaker 3: He's like, we didn't realize the power of prompting that 814 00:35:49,719 --> 00:35:51,960 Speaker 3: A lot of the AI researchers, if you go five 815 00:35:52,000 --> 00:35:54,640 Speaker 3: years ago, four years ago, three years ago, they were 816 00:35:54,680 --> 00:35:56,839 Speaker 3: just like going to make a better model with more parameters, 817 00:35:56,920 --> 00:35:59,960 Speaker 3: better training data. And I think one of the breakthrough 818 00:36:00,120 --> 00:36:02,520 Speaker 3: that happened both with GPT four and chat GPT to 819 00:36:02,560 --> 00:36:05,400 Speaker 3: some degree one was a chat interface that unlocked a 820 00:36:05,400 --> 00:36:08,080 Speaker 3: lot and then the other was that you can actually 821 00:36:08,200 --> 00:36:10,520 Speaker 3: and GPT four in particular was very steerable, which means 822 00:36:10,560 --> 00:36:12,480 Speaker 3: you could prompt it and have it take on personas 823 00:36:12,480 --> 00:36:14,800 Speaker 3: as a tutor or as a poet, or as a 824 00:36:14,880 --> 00:36:16,839 Speaker 3: rapper or whatever you want to make it do. And 825 00:36:17,080 --> 00:36:19,680 Speaker 3: they I don't think the researchers had the time to 826 00:36:19,760 --> 00:36:21,920 Speaker 3: even sit there and prompt it. So it is amazing 827 00:36:21,960 --> 00:36:24,600 Speaker 3: even they didn't realize how far you could get with 828 00:36:24,640 --> 00:36:27,960 Speaker 3: something like that five years in the future. You know, 829 00:36:28,520 --> 00:36:30,680 Speaker 3: there's some people are arguing that, you know, we're hitting 830 00:36:30,680 --> 00:36:33,279 Speaker 3: some type of an asymptote. You know, the jump from 831 00:36:33,320 --> 00:36:36,759 Speaker 3: GPT three to GPT four was massive. The jump from 832 00:36:36,800 --> 00:36:40,719 Speaker 3: GPT four now two years later to GPT four Omni 833 00:36:41,320 --> 00:36:44,719 Speaker 3: one reasoning model three or whatever, or Gemini three or 834 00:36:45,400 --> 00:36:46,480 Speaker 3: Claude three point five. 835 00:36:46,920 --> 00:36:47,520 Speaker 2: It's better. 836 00:36:47,640 --> 00:36:50,520 Speaker 3: I mean, the hallucinations are way down, the math errors 837 00:36:50,520 --> 00:36:50,960 Speaker 3: the way down. 838 00:36:51,080 --> 00:36:52,560 Speaker 2: It is getting better at reasoning. 839 00:36:52,239 --> 00:36:54,239 Speaker 3: But it's not as much of a you know, we 840 00:36:54,280 --> 00:36:56,480 Speaker 3: thought GPT five would be around by now, and that 841 00:36:56,680 --> 00:37:00,000 Speaker 3: you know, maybe it would be approaching you know, something 842 00:37:00,120 --> 00:37:03,839 Speaker 3: even more mind blowing or even transcending human intelligence that 843 00:37:03,920 --> 00:37:06,640 Speaker 3: hasn't happened yet. So I think there's two scenarios. I 844 00:37:06,640 --> 00:37:08,080 Speaker 3: think there's a world where in the next five years 845 00:37:08,120 --> 00:37:10,880 Speaker 3: we do and then that's a wild world. I think 846 00:37:10,920 --> 00:37:13,360 Speaker 3: you're going to see massive acceleration. I mean in your world, 847 00:37:13,400 --> 00:37:17,919 Speaker 3: understanding the brain, understanding complex systems, drug developments. 848 00:37:17,960 --> 00:37:18,400 Speaker 2: Science. 849 00:37:18,520 --> 00:37:20,759 Speaker 3: I've heard people say we're gonna have one hundred years 850 00:37:20,760 --> 00:37:23,040 Speaker 3: of science breakthroughs in the next ten I don't know. 851 00:37:23,440 --> 00:37:25,799 Speaker 2: I don't know. I think that's at least true. 852 00:37:25,880 --> 00:37:28,320 Speaker 1: I've actually made the argument that A is going to 853 00:37:28,440 --> 00:37:32,120 Speaker 1: help in science in several ways. The main way it's 854 00:37:32,160 --> 00:37:35,040 Speaker 1: going to do it is simply by pulling together facts 855 00:37:35,360 --> 00:37:37,319 Speaker 1: that I couldn't possibly know because I couldn't have read 856 00:37:37,320 --> 00:37:40,160 Speaker 1: all those journal articles. But the thing I'm really looking 857 00:37:40,160 --> 00:37:45,839 Speaker 1: forward to now, which current lms cannot do, is scientific 858 00:37:45,920 --> 00:37:49,920 Speaker 1: discovery of the type that humans do, which is extrapolating 859 00:37:49,960 --> 00:37:52,759 Speaker 1: thinking of a new framework, saying what if I were 860 00:37:52,800 --> 00:37:55,120 Speaker 1: writing on that photon of light, what would the world 861 00:37:55,200 --> 00:37:58,360 Speaker 1: look like? Oh? That leads me to special theory of relativity. 862 00:37:58,600 --> 00:38:02,839 Speaker 1: That's the kind of thing the current llms are incredible 863 00:38:02,880 --> 00:38:06,879 Speaker 1: at interpolating, but not at extrapolating. So that's what I'm 864 00:38:06,920 --> 00:38:09,319 Speaker 1: looking forward to in science. But actually, let me ask 865 00:38:09,360 --> 00:38:12,720 Speaker 1: you a question on this, And actually you mentioned hallucinations 866 00:38:12,719 --> 00:38:15,520 Speaker 1: a moment ago, and I always think about what I 867 00:38:15,680 --> 00:38:18,920 Speaker 1: like about hallucinations in MS, which is that you can 868 00:38:18,960 --> 00:38:22,919 Speaker 1: also think of it as hypotheses, which is to say, 869 00:38:23,360 --> 00:38:25,240 Speaker 1: and this is what scientists do all the time. 870 00:38:25,080 --> 00:38:26,359 Speaker 2: Is hey, what if? What if? 871 00:38:26,400 --> 00:38:28,200 Speaker 1: You know, we're trying things out all the time. Most 872 00:38:28,200 --> 00:38:30,440 Speaker 1: of them don't lead anywhere. Occasionally you can build a 873 00:38:30,480 --> 00:38:32,279 Speaker 1: bridge to that and say, oh my gosh, I just 874 00:38:32,320 --> 00:38:33,160 Speaker 1: made real progress. 875 00:38:33,719 --> 00:38:36,120 Speaker 2: So what do you think is the effect. 876 00:38:35,960 --> 00:38:41,280 Speaker 1: Of AIS in education on creativity in students. 877 00:38:41,880 --> 00:38:43,520 Speaker 3: I read a lot about it in my book Brave 878 00:38:43,600 --> 00:38:46,279 Speaker 3: New Words, and I think it's going to be a 879 00:38:46,360 --> 00:38:48,759 Speaker 3: creativity amplifier. You know, some people would argue like, oh, 880 00:38:48,800 --> 00:38:50,719 Speaker 3: if this thing can create a lazy student's just not 881 00:38:50,719 --> 00:38:53,000 Speaker 3: going to create themselves. And look, lazy students are going 882 00:38:53,000 --> 00:38:54,480 Speaker 3: to figure out ways to be lazy. There's ways to 883 00:38:54,480 --> 00:38:57,040 Speaker 3: police them. We're working on that too. But I always 884 00:38:57,040 --> 00:38:59,680 Speaker 3: point out the most creative moments in my life have 885 00:38:59,760 --> 00:39:01,640 Speaker 3: been when I'm in a room with other creative people. 886 00:39:01,880 --> 00:39:03,600 Speaker 3: If you look at all the creative moments in history, 887 00:39:03,640 --> 00:39:06,120 Speaker 3: you know, Florence during the Renaissance, or some of the 888 00:39:07,040 --> 00:39:09,640 Speaker 3: you know beat nicks in you know, different cities in 889 00:39:09,640 --> 00:39:11,040 Speaker 3: the world, or Silicon Valley. 890 00:39:11,120 --> 00:39:12,520 Speaker 2: Right now, it's there. 891 00:39:12,680 --> 00:39:16,760 Speaker 3: There is this effect of people meeting other creative people 892 00:39:16,800 --> 00:39:19,480 Speaker 3: and then collaborating with them, and the power of AI 893 00:39:19,760 --> 00:39:22,040 Speaker 3: is you know, like I remember being in middle school 894 00:39:22,040 --> 00:39:24,799 Speaker 3: in high school and I was a curious kid, but 895 00:39:24,840 --> 00:39:27,800 Speaker 3: I had to suppress that curiosity in the classroom otherwise 896 00:39:27,800 --> 00:39:29,719 Speaker 3: I would get beat up. But I had no one 897 00:39:29,719 --> 00:39:31,520 Speaker 3: to talk about this stuff with, and you just kind 898 00:39:31,520 --> 00:39:33,640 Speaker 3: of sit been by yourself. It's like, I wonder, I 899 00:39:33,680 --> 00:39:35,239 Speaker 3: wonder if this is different, and I wonder if I 900 00:39:35,280 --> 00:39:39,120 Speaker 3: could do this. But now, if young sal had had 901 00:39:39,200 --> 00:39:41,520 Speaker 3: chat GBT or Gemini or something, I. 902 00:39:41,440 --> 00:39:42,880 Speaker 2: Could have played around. I was like, well, what do 903 00:39:42,920 --> 00:39:44,279 Speaker 2: you mean by that? Or how does this? 904 00:39:44,680 --> 00:39:46,640 Speaker 3: And I do think you know, to your point, some 905 00:39:46,680 --> 00:39:50,719 Speaker 3: of the big breakthroughs are naive thinking, and some of 906 00:39:50,719 --> 00:39:53,200 Speaker 3: the best naive thinking is coming from people who haven't 907 00:39:53,200 --> 00:39:56,480 Speaker 3: been indoctrinated yet. And every couple of weeks I do 908 00:39:56,480 --> 00:39:59,000 Speaker 3: get an email from a student saying, Oh, I watched 909 00:39:59,000 --> 00:40:01,240 Speaker 3: your video on this, but I have a new theory 910 00:40:01,960 --> 00:40:05,560 Speaker 3: that disproves you know Newton, And I'm like, okay, whatever, whatever, 911 00:40:05,719 --> 00:40:07,879 Speaker 3: But you know, one out of every thousand of those 912 00:40:07,920 --> 00:40:10,279 Speaker 3: kids is going to be right if they can start 913 00:40:10,320 --> 00:40:13,400 Speaker 3: to validate or fine tune their ideas with an AI 914 00:40:13,520 --> 00:40:17,319 Speaker 3: that could even bring what's already known out there. One 915 00:40:17,360 --> 00:40:19,200 Speaker 3: of them is like, wait, I think I'm onto something, 916 00:40:19,640 --> 00:40:21,560 Speaker 3: and maybe the AI says, you know, you might be 917 00:40:21,680 --> 00:40:41,440 Speaker 3: onto something that might reconcile gravity with quantum mechanics. 918 00:40:43,680 --> 00:40:46,080 Speaker 1: It feels to me like the two most important things 919 00:40:46,680 --> 00:40:49,440 Speaker 1: for educators is, you know, roots and wings. This was 920 00:40:49,480 --> 00:40:52,919 Speaker 1: something Gerta said two hundred and fifty years ago that 921 00:40:52,560 --> 00:40:54,960 Speaker 1: that's our job is to give kids roots and wings, 922 00:40:55,040 --> 00:40:58,279 Speaker 1: meaning all the content knowledge as you mentioned, and the 923 00:40:58,280 --> 00:41:02,080 Speaker 1: wings being the creativity part. So if students have a 924 00:41:02,239 --> 00:41:05,319 Speaker 1: partner with which they can do it, great, that makes 925 00:41:05,320 --> 00:41:08,279 Speaker 1: things better. How do we make sure that kids are 926 00:41:08,680 --> 00:41:10,880 Speaker 1: using it as a partner as opposed to being the 927 00:41:10,960 --> 00:41:13,480 Speaker 1: lazy kid and just letting let me do it. 928 00:41:13,880 --> 00:41:15,879 Speaker 2: Yeah. I wrote about a whole in my book. 929 00:41:15,880 --> 00:41:18,880 Speaker 3: I wrote a whole chapter about cheating, and I always 930 00:41:18,920 --> 00:41:20,960 Speaker 3: like to start like, you know, everyone's concerned about cheating 931 00:41:20,960 --> 00:41:23,120 Speaker 3: because of chat GEPT and it can cheat. I was like, 932 00:41:23,120 --> 00:41:25,120 Speaker 3: but what was the state of cheating before, and it 933 00:41:25,160 --> 00:41:26,839 Speaker 3: turns out there's been a lot of cheating, in fact, 934 00:41:26,920 --> 00:41:29,960 Speaker 3: even since you and I were in school. Unfortunately cheating, 935 00:41:30,040 --> 00:41:32,799 Speaker 3: you know, I think the internet, there's services online that 936 00:41:33,040 --> 00:41:35,640 Speaker 3: a human being will write your paper in Kenya, a 937 00:41:35,680 --> 00:41:37,760 Speaker 3: PhD in English will write your paper for five dollars 938 00:41:37,800 --> 00:41:40,440 Speaker 3: a page, write your term paper for you. And that's 939 00:41:40,680 --> 00:41:43,480 Speaker 3: arguably and that's been around for twenty years. So and 940 00:41:43,760 --> 00:41:47,880 Speaker 3: sorority A sororities sisters and fraternity brothers are you know, sharing, 941 00:41:48,280 --> 00:41:51,440 Speaker 3: So that's always been there and so our solution to 942 00:41:51,480 --> 00:41:53,640 Speaker 3: it is, yeah, you're right, chat GPT could be used. 943 00:41:53,680 --> 00:41:55,200 Speaker 2: In fact, I'm sure people are using it. 944 00:41:55,560 --> 00:41:57,440 Speaker 3: And anyone who says that there's tools that can detect 945 00:41:57,440 --> 00:41:59,640 Speaker 3: AI cheating is it snake oil. In fact, there's a 946 00:41:59,640 --> 00:42:02,640 Speaker 3: lot of FA accusations happening right now, especially in higher education. 947 00:42:03,160 --> 00:42:05,640 Speaker 3: But I think the real solution is and it doesn't 948 00:42:05,640 --> 00:42:07,920 Speaker 3: just solve the cheating problem from AI. I think it 949 00:42:08,120 --> 00:42:10,720 Speaker 3: solved the cheating problem generally, but it can also support 950 00:42:10,760 --> 00:42:14,880 Speaker 3: students and teachers better is have the students have the 951 00:42:14,920 --> 00:42:17,440 Speaker 3: teacher signed through the AI, have the students work with 952 00:42:17,480 --> 00:42:19,520 Speaker 3: the AI on the paper, and then have the AI 953 00:42:19,600 --> 00:42:22,839 Speaker 3: make the entire process transparent to the professor or to 954 00:42:22,880 --> 00:42:26,520 Speaker 3: the teacher, because then and we're building this. But then 955 00:42:26,680 --> 00:42:29,400 Speaker 3: if the AI could say, yes, professor, we worked on 956 00:42:29,440 --> 00:42:31,440 Speaker 3: this for four hours. David had some trouble coming up 957 00:42:31,440 --> 00:42:33,360 Speaker 3: with a thesis statement. This is where we worked. You 958 00:42:33,440 --> 00:42:35,120 Speaker 3: can see two hours into it. We had a lot 959 00:42:35,160 --> 00:42:36,920 Speaker 3: of back and forth on this part. You can see 960 00:42:36,920 --> 00:42:39,359 Speaker 3: his edit history, but you know, three hours into it, 961 00:42:39,400 --> 00:42:41,600 Speaker 3: he just copy and pasted this paragraph into this. I 962 00:42:41,640 --> 00:42:43,359 Speaker 3: do not know where it comes from. It actually doesn't 963 00:42:43,360 --> 00:42:46,279 Speaker 3: look like his writing. We should look into this. Don't 964 00:42:46,280 --> 00:42:48,880 Speaker 3: accuse him, ask him about that paragraph. I think that 965 00:42:48,880 --> 00:42:51,280 Speaker 3: could be healthy. Or if the whole thing just appears 966 00:42:51,680 --> 00:42:54,080 Speaker 3: or it just comes linearly on, and you know, it 967 00:42:54,080 --> 00:42:56,160 Speaker 3: doesn't look like something that someone worked on, then yeah, 968 00:42:56,160 --> 00:42:58,239 Speaker 3: that's the red flag. But that would have flagged even 969 00:42:58,320 --> 00:43:01,600 Speaker 3: the situation where you hired someone in Africa to write 970 00:43:01,640 --> 00:43:03,000 Speaker 3: your paper for you and. 971 00:43:02,960 --> 00:43:05,640 Speaker 2: It can quiz you on it. Yes, that's excellent. 972 00:43:06,280 --> 00:43:08,680 Speaker 1: One question just to push back on that is will 973 00:43:08,719 --> 00:43:12,720 Speaker 1: teachers want to look through the scrolls of data about 974 00:43:12,920 --> 00:43:14,440 Speaker 1: this paragraph appeared and so on? 975 00:43:14,560 --> 00:43:16,440 Speaker 2: Or is that too much work for them? Well, it 976 00:43:16,520 --> 00:43:16,759 Speaker 2: is true. 977 00:43:16,840 --> 00:43:18,520 Speaker 3: Well, That's why it's useful for the AI, because the 978 00:43:18,520 --> 00:43:21,920 Speaker 3: AI could say so right now, our product manager, who 979 00:43:21,960 --> 00:43:23,360 Speaker 3: leads what we call writing coach, she used to be 980 00:43:23,400 --> 00:43:25,680 Speaker 3: an English teacher, and she used to have one hundred 981 00:43:25,800 --> 00:43:30,440 Speaker 3: seventh grade students on her docket. She assigns one paper 982 00:43:31,320 --> 00:43:34,200 Speaker 3: to grade. Those hundred papers would take her seventeen hours. 983 00:43:34,960 --> 00:43:38,000 Speaker 3: That is not a great like that seventeen hours above 984 00:43:38,000 --> 00:43:40,799 Speaker 3: and beyond being at the school for forty like it's incredible, 985 00:43:41,760 --> 00:43:44,040 Speaker 3: and teachers are doing this right now. So imagine this 986 00:43:44,120 --> 00:43:47,600 Speaker 3: new world where those hundred papers are done. And even 987 00:43:47,600 --> 00:43:49,440 Speaker 3: those papers you don't know if Chad Gpt did them, 988 00:43:49,440 --> 00:43:52,320 Speaker 3: you didn't know what happened. Now imagine this new world 989 00:43:52,560 --> 00:43:55,399 Speaker 3: the papers are all done. The AI, first of all, 990 00:43:55,520 --> 00:43:58,560 Speaker 3: can say, you know her name is Sarah Sarah, based 991 00:43:58,600 --> 00:44:01,800 Speaker 3: on our rubric we constructed together, here's my preliminary grade 992 00:44:02,040 --> 00:44:05,239 Speaker 3: and my justification of why. But you should review it. 993 00:44:05,800 --> 00:44:08,319 Speaker 3: So what used to take seventeen hours can now maybe 994 00:44:08,320 --> 00:44:10,759 Speaker 3: take you two hours. You still should review it, but 995 00:44:10,840 --> 00:44:12,360 Speaker 3: like oh this, and in a lot of ways, it 996 00:44:12,400 --> 00:44:14,840 Speaker 3: might be far more consistent because if you're grading one 997 00:44:14,920 --> 00:44:19,200 Speaker 3: hundred seventh grade papers on the great Gatsby that ninetieth paper. 998 00:44:19,480 --> 00:44:21,320 Speaker 3: You know you're not going to be consistent. I wouldn't 999 00:44:21,360 --> 00:44:24,920 Speaker 3: be consistent one hundred percent. But also yes, and I'm 1000 00:44:24,960 --> 00:44:26,680 Speaker 3: confident that this is David's work, et cetera. 1001 00:44:26,800 --> 00:44:27,279 Speaker 2: It's just giving. 1002 00:44:27,360 --> 00:44:29,720 Speaker 3: It can give the synopsis, but then Sarah can ask questions. 1003 00:44:29,719 --> 00:44:31,520 Speaker 3: She's like, tell me more. Are there some themes you see? 1004 00:44:31,640 --> 00:44:33,160 Speaker 3: But yeah, it turns out a lot of your students 1005 00:44:33,200 --> 00:44:35,960 Speaker 3: are having trouble with thesis statements. So Sarah doesn't have 1006 00:44:35,960 --> 00:44:39,520 Speaker 3: to read every transcript that's not But once again, this 1007 00:44:39,600 --> 00:44:41,640 Speaker 3: acts as a teaching assistant and it can do all 1008 00:44:41,640 --> 00:44:44,040 Speaker 3: of that grunt work, but give Sarah way way more 1009 00:44:44,080 --> 00:44:48,160 Speaker 3: insights on where the students are, streamline the grading, and 1010 00:44:48,400 --> 00:44:50,160 Speaker 3: undermine any kind of cheating that might happen. 1011 00:44:50,560 --> 00:44:50,959 Speaker 2: I see. 1012 00:44:51,000 --> 00:44:53,680 Speaker 1: And it's not only grading the individual papers, but it's 1013 00:44:53,680 --> 00:44:56,239 Speaker 1: also giving her feedback about how the group did. 1014 00:44:56,680 --> 00:44:57,759 Speaker 2: Yeah, yeah, we should do. 1015 00:44:57,920 --> 00:45:00,520 Speaker 3: Let's let's modify tomorrow's lesson plan to be focused on outlining, 1016 00:45:00,560 --> 00:45:02,080 Speaker 3: because that's where a lot of students struggled. 1017 00:45:02,280 --> 00:45:05,239 Speaker 1: In your experience, how are teachers currently feeling about AAR 1018 00:45:05,400 --> 00:45:08,279 Speaker 1: feeling excited? Are they feeling threatened? And what advice do 1019 00:45:08,280 --> 00:45:12,160 Speaker 1: you give to teachers. I've been pleasantly surprised. You know, 1020 00:45:12,200 --> 00:45:14,520 Speaker 1: when we started working on kon Mega, we had we 1021 00:45:14,560 --> 00:45:16,440 Speaker 1: had all these debates inside of our organization, people like 1022 00:45:16,520 --> 00:45:19,560 Speaker 1: it's communities for cheating, and had hallucinates and and what 1023 00:45:19,680 --> 00:45:21,319 Speaker 1: I what I told the team at the time is like, look, 1024 00:45:21,360 --> 00:45:23,319 Speaker 1: we we got to just this is going to be 1025 00:45:23,320 --> 00:45:23,560 Speaker 1: a thing. 1026 00:45:23,560 --> 00:45:25,239 Speaker 3: It's going to be it's going to improve. We have 1027 00:45:25,360 --> 00:45:27,120 Speaker 3: these are real concerns, but we have to turn them 1028 00:45:27,160 --> 00:45:29,080 Speaker 3: into features. So we said, Okay, we're gonna create oversight. 1029 00:45:29,080 --> 00:45:32,040 Speaker 3: We're going to moderate what students are doing. We're gonna 1030 00:45:32,840 --> 00:45:34,319 Speaker 3: cot Migo's not going to give you the answer. It's 1031 00:45:34,320 --> 00:45:36,480 Speaker 3: going to be socratic. It's going to ask questions. But 1032 00:45:36,560 --> 00:45:38,919 Speaker 3: I was still afraid that when we launched, like there's 1033 00:45:38,960 --> 00:45:40,879 Speaker 3: there's gonna be a lot of blowback. Con Academy's going 1034 00:45:40,880 --> 00:45:43,319 Speaker 3: all in on AI. It's not perfect, et cetera. We 1035 00:45:43,320 --> 00:45:45,240 Speaker 3: were working on that, and then all of a sudden 1036 00:45:45,280 --> 00:45:47,520 Speaker 3: and we were all under a non disclosure agreement with 1037 00:45:47,600 --> 00:45:50,839 Speaker 3: chat GPT, with open Ai all lawyered up, and and 1038 00:45:50,880 --> 00:45:52,879 Speaker 3: then and we were supposed to launch March of twenty 1039 00:45:52,920 --> 00:45:56,560 Speaker 3: twenty three, November twenty twenty two, the last day of November. 1040 00:45:56,239 --> 00:45:58,719 Speaker 2: Chat GPT comes out, and I remember, like, what's going on. 1041 00:45:58,719 --> 00:46:00,720 Speaker 3: You guys said you were going to launch anything until March, 1042 00:46:01,160 --> 00:46:03,120 Speaker 3: and then I slacked Greg Brock and he's like, oh no, 1043 00:46:03,160 --> 00:46:04,040 Speaker 3: we all we did is put it. 1044 00:46:04,040 --> 00:46:06,040 Speaker 2: We launched all these apps on our old model. 1045 00:46:06,719 --> 00:46:09,200 Speaker 3: Chat GPT was one of them, just a chat interface 1046 00:46:09,200 --> 00:46:12,719 Speaker 3: on GPT three point five. We too are surprised. We 1047 00:46:13,280 --> 00:46:15,120 Speaker 3: didn't think anyone would care until GPT. 1048 00:46:14,880 --> 00:46:15,760 Speaker 2: Four came out. 1049 00:46:15,920 --> 00:46:17,960 Speaker 3: We were really wired because that was the moment where 1050 00:46:17,960 --> 00:46:19,520 Speaker 3: everyone says, oh my god, this thing has a lot 1051 00:46:19,560 --> 00:46:22,320 Speaker 3: of errors, it's bad at math, and it's a cheating tool. 1052 00:46:22,520 --> 00:46:25,560 Speaker 3: And then you saw these school districts start to ban it, etc. 1053 00:46:25,880 --> 00:46:27,359 Speaker 3: And we're like, oh no, we're working on this thing. 1054 00:46:27,360 --> 00:46:29,600 Speaker 3: We're gonna release in March, and everyone's already assuming that 1055 00:46:29,680 --> 00:46:32,040 Speaker 3: this is evil. It ended up that that was actually 1056 00:46:32,080 --> 00:46:35,040 Speaker 3: a blessing because by the time March came around four 1057 00:46:35,040 --> 00:46:38,840 Speaker 3: months later, the school system actually had said or started 1058 00:46:38,840 --> 00:46:41,280 Speaker 3: to think, well, you know what, this does have issues, 1059 00:46:41,960 --> 00:46:44,239 Speaker 3: but if only someone were to mitigate the risks, but 1060 00:46:44,320 --> 00:46:46,440 Speaker 3: this is going to be part of kids' lives and 1061 00:46:46,560 --> 00:46:48,960 Speaker 3: it couldn't theory be used to teach them? What if 1062 00:46:49,000 --> 00:46:50,319 Speaker 3: we could use it that way? And so when we 1063 00:46:50,400 --> 00:46:53,799 Speaker 3: launched Conmego, it had positive reaction. And I think what 1064 00:46:53,840 --> 00:46:56,279 Speaker 3: we're seeing from teachers is for some teachers there is 1065 00:46:56,280 --> 00:46:58,839 Speaker 3: a little trepidation, but as soon as we make clear, like, look, 1066 00:46:59,000 --> 00:47:01,160 Speaker 3: if I harted teaching sistant for every one of you, you 1067 00:47:01,080 --> 00:47:03,359 Speaker 3: would love that and you wouldn't feel threatened no matter 1068 00:47:03,400 --> 00:47:05,879 Speaker 3: how amazing that teaching assistant is. As long as you're 1069 00:47:05,880 --> 00:47:08,200 Speaker 3: in charge, that's what this is. And then Amani like, 1070 00:47:08,239 --> 00:47:11,080 Speaker 3: oh wow, yeah, this is really a gift. But and 1071 00:47:11,239 --> 00:47:13,279 Speaker 3: the other benefit is teachers are immediately seeing how it 1072 00:47:13,320 --> 00:47:15,960 Speaker 3: benefits their lives. I give that example of an English 1073 00:47:15,960 --> 00:47:19,440 Speaker 3: teacher having to spend seventeen hours grading one assignment, or 1074 00:47:19,480 --> 00:47:21,960 Speaker 3: if you're spending ten hours a week lesson planning or 1075 00:47:21,960 --> 00:47:24,600 Speaker 3: writing progress reports, and if you can shorten that by 1076 00:47:24,600 --> 00:47:27,120 Speaker 3: a factor of five or a factor of ten, that's 1077 00:47:27,160 --> 00:47:28,720 Speaker 3: an immediate benefit for the teacher. 1078 00:47:28,920 --> 00:47:30,360 Speaker 2: And it's a benefit for the teacher. 1079 00:47:30,560 --> 00:47:32,440 Speaker 3: If you have a classroom of thirty kids and they're 1080 00:47:32,440 --> 00:47:35,800 Speaker 3: all at different levels, you can't replicate yourself to address 1081 00:47:35,800 --> 00:47:37,719 Speaker 3: every one of their needs. So if more of those 1082 00:47:37,760 --> 00:47:40,320 Speaker 3: kids are having their gaps filled in and you're getting 1083 00:47:40,320 --> 00:47:42,600 Speaker 3: that information and then you can up level the lessons 1084 00:47:42,600 --> 00:47:45,000 Speaker 3: and it takes you less time to prepare them. That's 1085 00:47:45,000 --> 00:47:48,359 Speaker 3: starting to get to, you know, better and better from 1086 00:47:48,360 --> 00:47:49,239 Speaker 3: the teacher's point of view. 1087 00:47:49,680 --> 00:47:52,520 Speaker 1: What is the most surprising way that teachers or students 1088 00:47:52,560 --> 00:47:53,680 Speaker 1: have used your platform? 1089 00:47:54,280 --> 00:47:57,759 Speaker 3: I think early on I write about this in Brave 1090 00:47:57,840 --> 00:48:00,560 Speaker 3: New Words. This was I saw my opened the book 1091 00:48:00,560 --> 00:48:02,600 Speaker 3: this way when my daughter. This was in the early 1092 00:48:02,680 --> 00:48:05,000 Speaker 3: days before we had launched con Migo. We had access 1093 00:48:05,040 --> 00:48:08,080 Speaker 3: to GPT four no one else did. I had prompted 1094 00:48:08,120 --> 00:48:10,160 Speaker 3: it to just kind of co write a story with her, 1095 00:48:10,640 --> 00:48:12,720 Speaker 3: and it was a story about the social media influencer 1096 00:48:12,760 --> 00:48:15,239 Speaker 3: Samantha who's stuck on a desert island and she was 1097 00:48:15,280 --> 00:48:17,600 Speaker 3: having like a panic attack because she couldn't share how 1098 00:48:17,640 --> 00:48:21,200 Speaker 3: beautiful it was with her friends. And then my daughter said, 1099 00:48:21,200 --> 00:48:24,799 Speaker 3: can I talk to Samantha. I'm like, I guess you could, 1100 00:48:24,880 --> 00:48:26,320 Speaker 3: and so she said, I'd like to talk to Samantha, 1101 00:48:26,320 --> 00:48:30,560 Speaker 3: and then the aimedia says, hey, Samantha, here, I don't 1102 00:48:30,560 --> 00:48:32,440 Speaker 3: know what to do. I can't share this, and my 1103 00:48:32,520 --> 00:48:35,600 Speaker 3: daughter's like, it's okay, you know, just enjoy the beauty 1104 00:48:35,760 --> 00:48:36,160 Speaker 3: and like. 1105 00:48:36,480 --> 00:48:37,839 Speaker 2: And that was the first one. I was like, this 1106 00:48:37,880 --> 00:48:38,560 Speaker 2: is science fiction. 1107 00:48:38,680 --> 00:48:40,840 Speaker 3: Like, my daughter is not only writing a story with 1108 00:48:40,880 --> 00:48:43,520 Speaker 3: an AI, but now she can actually talk to a 1109 00:48:43,600 --> 00:48:47,799 Speaker 3: character that they constructed together in her story. And that 1110 00:48:47,960 --> 00:48:50,360 Speaker 3: was some of the idea behind Now we have activities 1111 00:48:50,400 --> 00:48:53,319 Speaker 3: on con Migo where you can talk to AI simulations 1112 00:48:53,360 --> 00:48:57,160 Speaker 3: of literary figures or historical characters. And you know, we 1113 00:48:57,560 --> 00:48:59,200 Speaker 3: have something called con World School, which is a virtual 1114 00:48:59,280 --> 00:49:02,000 Speaker 3: high school, and one of there's a student in India 1115 00:49:02,440 --> 00:49:04,600 Speaker 3: and she was starting The Great Gatsby. She had a huge, 1116 00:49:04,680 --> 00:49:07,520 Speaker 3: long conversation with Jay Gatsby AI simulation of Jay Gatsby, 1117 00:49:07,560 --> 00:49:09,640 Speaker 3: And you know, she told me the whole conversation. She's like, 1118 00:49:09,680 --> 00:49:11,439 Speaker 3: and I apologize for taking up all of his time. 1119 00:49:11,719 --> 00:49:13,799 Speaker 3: And she knew that he's an AI. It's not like 1120 00:49:13,880 --> 00:49:16,680 Speaker 3: she's you know, but and it was a pretty meaningful 1121 00:49:16,719 --> 00:49:20,279 Speaker 3: conversation about you know, goals and aspirations and never being 1122 00:49:20,280 --> 00:49:22,880 Speaker 3: able to reach them, et cetera. You know, I'm a 1123 00:49:22,880 --> 00:49:25,520 Speaker 3: Star trek nerd, especially you know, the next generation of 1124 00:49:25,520 --> 00:49:27,200 Speaker 3: the Holidack. And I used to think that was the 1125 00:49:27,239 --> 00:49:29,520 Speaker 3: most fake part of Star Trek, Like even in the 1126 00:49:29,560 --> 00:49:31,480 Speaker 3: twenty fourth century. Surely we're not going to be able 1127 00:49:31,480 --> 00:49:35,319 Speaker 3: to have like conversations with Albert Einstein. But literally, that's 1128 00:49:35,400 --> 00:49:38,200 Speaker 3: already hap there, that's already happening, and it's only going 1129 00:49:38,239 --> 00:49:39,000 Speaker 3: to get better and better. 1130 00:49:43,640 --> 00:49:45,840 Speaker 1: That was my interview with Saul Khan, founder of the 1131 00:49:45,880 --> 00:49:48,640 Speaker 1: Kahn Academy. One of the things I thought was interesting 1132 00:49:48,680 --> 00:49:51,719 Speaker 1: is that he's in favor of teaching wrote knowledge. Sometimes 1133 00:49:51,760 --> 00:49:54,600 Speaker 1: he argues that it is to some degree a really 1134 00:49:54,640 --> 00:49:57,840 Speaker 1: important factor. But of all the things we covered today, 1135 00:49:57,880 --> 00:50:00,560 Speaker 1: I have to say that my absolute favorite is the 1136 00:50:00,680 --> 00:50:05,920 Speaker 1: project of learning critical thinking by debating with an AI, 1137 00:50:06,239 --> 00:50:09,080 Speaker 1: one that is friendly but firm, and the idea that 1138 00:50:09,080 --> 00:50:11,360 Speaker 1: we can start doing this right away. We can start 1139 00:50:11,440 --> 00:50:14,720 Speaker 1: getting our students to take a side on some issue 1140 00:50:14,960 --> 00:50:19,120 Speaker 1: and put forward logical arguments about it, and get graded 1141 00:50:19,200 --> 00:50:22,480 Speaker 1: on the quality of their arguments, and then on their 1142 00:50:22,600 --> 00:50:26,680 Speaker 1: ability to represent the argument of the other side, and 1143 00:50:26,760 --> 00:50:30,279 Speaker 1: then to switch sides and do it all again. That 1144 00:50:30,360 --> 00:50:33,240 Speaker 1: feels to me like a place where AI can fill 1145 00:50:33,280 --> 00:50:36,400 Speaker 1: a niche that is not currently filled, and it can 1146 00:50:36,480 --> 00:50:40,680 Speaker 1: do this with the loving patience that we would devoutly 1147 00:50:40,719 --> 00:50:44,320 Speaker 1: wish for when arguing with our classmates. Or our children 1148 00:50:44,400 --> 00:50:48,200 Speaker 1: or our students. The AI never gets tired or bored 1149 00:50:48,440 --> 00:50:52,279 Speaker 1: or offended. So in the end, as we start each 1150 00:50:52,320 --> 00:50:55,920 Speaker 1: school year in a world that changes faster and faster, 1151 00:50:56,480 --> 00:51:00,560 Speaker 1: I have high hopes that will continue to discover new 1152 00:51:00,600 --> 00:51:04,680 Speaker 1: ways to teach our students things that last, that we 1153 00:51:04,800 --> 00:51:07,400 Speaker 1: will give them roots and wings. 1154 00:51:07,719 --> 00:51:09,279 Speaker 2: And if we can use AI. 1155 00:51:09,360 --> 00:51:13,480 Speaker 1: To help students sharpen their reasoning and engage in deeper 1156 00:51:13,520 --> 00:51:17,960 Speaker 1: conversations and see the world from multiple perspectives, then we're 1157 00:51:17,960 --> 00:51:21,719 Speaker 1: not just preparing them to take the standardized tests. We 1158 00:51:21,800 --> 00:51:26,560 Speaker 1: are preparing them for life. The future of education isn't 1159 00:51:26,600 --> 00:51:29,759 Speaker 1: just about what we learn, but how we learn, and 1160 00:51:29,800 --> 00:51:33,000 Speaker 1: with the right technology, I have no doubt that we 1161 00:51:33,080 --> 00:51:42,520 Speaker 1: can make that future and our students quite bright. Go 1162 00:51:42,560 --> 00:51:45,600 Speaker 1: to Eagleman dot com slash podcasts for more information and 1163 00:51:45,640 --> 00:51:47,000 Speaker 1: to find further reading. 1164 00:51:47,480 --> 00:51:48,800 Speaker 2: Send me an email. 1165 00:51:48,480 --> 00:51:51,759 Speaker 1: At podcasts at eagleman dot com with questions or discussion, 1166 00:51:52,120 --> 00:51:55,319 Speaker 1: and check out and subscribe to Intercosmos on YouTube for 1167 00:51:55,480 --> 00:51:59,439 Speaker 1: videos of each episode and to leave comments until next time. 1168 00:51:59,560 --> 00:52:02,560 Speaker 1: I'm Dave Eagelman and this is inner Cosmos.