1 00:00:00,160 --> 00:00:02,960 Speaker 1: Hey everyone, it's Robert and Joe here. Today we've got 2 00:00:02,960 --> 00:00:05,240 Speaker 1: something a little bit different to share with you. It's 3 00:00:05,280 --> 00:00:09,400 Speaker 1: a new season of the Smart Talks with IBM podcast series. 4 00:00:09,720 --> 00:00:13,520 Speaker 2: This season on Smart Talks with IBM, Malcolm Gladwell is back, 5 00:00:13,520 --> 00:00:15,760 Speaker 2: and this time he's taking the show on the road. 6 00:00:15,880 --> 00:00:19,160 Speaker 2: Malcolm is stepping outside the studio to explore how IBM 7 00:00:19,239 --> 00:00:22,880 Speaker 2: clients are using artificial intelligence to solve real world challenges 8 00:00:23,079 --> 00:00:25,440 Speaker 2: and transform the way they do business. 9 00:00:25,480 --> 00:00:29,560 Speaker 1: From accelerating scientific breakthroughs to reimagining education. It's a fresh 10 00:00:29,600 --> 00:00:33,040 Speaker 1: look at innovation in action, where big ideas meet cutting 11 00:00:33,120 --> 00:00:33,960 Speaker 1: edge solutions. 12 00:00:34,320 --> 00:00:37,160 Speaker 2: You'll hear from industry leaders, creative thinkers, and of course 13 00:00:37,280 --> 00:00:40,920 Speaker 2: Malcolm Gladwell himself as he guides you through each story. 14 00:00:41,440 --> 00:00:44,680 Speaker 1: New episodes of Smart Talks with IBM drop every month 15 00:00:44,680 --> 00:00:48,279 Speaker 1: on the iHeartRadio app, Apple Podcasts, or wherever you get 16 00:00:48,320 --> 00:00:52,240 Speaker 1: your podcasts. Learn more at IBM dot com slash smart Talks. 17 00:00:52,880 --> 00:00:56,200 Speaker 1: This is a paid advertisement from IBM. 18 00:00:56,400 --> 00:00:59,440 Speaker 3: In the world of educational research, is a famous video 19 00:00:59,480 --> 00:01:02,680 Speaker 3: of a boy named Sean. I don't mean famous in 20 00:01:02,680 --> 00:01:04,679 Speaker 3: a sense that it has a million views on YouTube. 21 00:01:05,120 --> 00:01:07,080 Speaker 3: I mean that in the circle of people who think 22 00:01:07,120 --> 00:01:10,640 Speaker 3: about teaching and how to make teaching better. The video 23 00:01:10,680 --> 00:01:14,240 Speaker 3: has been written about in journal articles and shown over 24 00:01:14,280 --> 00:01:17,840 Speaker 3: and again in college classrooms. It's a ten minute clip 25 00:01:18,240 --> 00:01:21,520 Speaker 3: of a third grade class somewhere in Michigan. It was 26 00:01:21,560 --> 00:01:24,280 Speaker 3: filmed in January of nineteen ninety, so the video is 27 00:01:24,319 --> 00:01:27,959 Speaker 3: a bit grainy. The teacher's name is Deborah Lowenberg Ball. 28 00:01:28,319 --> 00:01:31,360 Speaker 3: She's a professor at Michigan State University who is part 29 00:01:31,360 --> 00:01:34,080 Speaker 3: of her research. Teaches a one hour math class at 30 00:01:34,120 --> 00:01:37,680 Speaker 3: a local elementary school on the day in question. Miss 31 00:01:37,760 --> 00:01:41,440 Speaker 3: Ball begins by asking her students about the previous day's lesson, 32 00:01:41,800 --> 00:01:44,600 Speaker 3: which was about even and odd numbers. 33 00:01:44,760 --> 00:01:47,000 Speaker 4: I would like to hear from as many people as possible, 34 00:01:47,400 --> 00:01:49,640 Speaker 4: what comments you had, reactions you had to be in 35 00:01:49,640 --> 00:01:50,560 Speaker 4: that meeting yesterday. 36 00:01:51,480 --> 00:01:54,320 Speaker 3: A little boy with black hair raises his hand. His 37 00:01:54,400 --> 00:01:55,000 Speaker 3: name is Sean. 38 00:01:55,760 --> 00:01:56,000 Speaker 4: Hello. 39 00:01:56,360 --> 00:01:59,120 Speaker 5: I don't have anything about the meeting yesterday. 40 00:01:59,160 --> 00:02:01,880 Speaker 4: That was Sean was. 41 00:02:01,840 --> 00:02:03,480 Speaker 3: Thinking about the number six. 42 00:02:04,120 --> 00:02:06,680 Speaker 5: So I was thinking that it's a it's an idd 43 00:02:06,800 --> 00:02:08,720 Speaker 5: it can be an odd number two because there could 44 00:02:08,760 --> 00:02:17,160 Speaker 5: be two, two, four, six, two, three twos and two 45 00:02:17,400 --> 00:02:26,480 Speaker 5: threes and add Antonisina two thinks make Italy two things. 46 00:02:26,120 --> 00:02:29,800 Speaker 3: And Sean doesn't understand what odd and even means. He 47 00:02:29,919 --> 00:02:32,160 Speaker 3: thinks that just because you can break down six in 48 00:02:32,240 --> 00:02:35,880 Speaker 3: an odd number of parts and an even number of parts, 49 00:02:35,919 --> 00:02:40,000 Speaker 3: that six must exist in some magical middle category. And 50 00:02:40,040 --> 00:02:42,800 Speaker 3: when you listen to the Sean videotape, you keep waiting 51 00:02:42,840 --> 00:02:46,080 Speaker 3: for the teacher to say, oh, no, Sean, you misunderstand. 52 00:02:46,680 --> 00:02:49,640 Speaker 3: But Deborah Ball doesn't do that. She never tells him 53 00:02:49,639 --> 00:02:54,799 Speaker 3: he's wrong. Instead, she simply asks him to explain his thinking. 54 00:02:55,200 --> 00:02:56,960 Speaker 2: And the two things that you put together to make 55 00:02:57,000 --> 00:02:57,760 Speaker 2: it were odd right? 56 00:02:58,440 --> 00:03:02,560 Speaker 6: Three and three are each old and thinks Roba so. 57 00:03:03,639 --> 00:03:07,359 Speaker 3: Two or even bald and asked the class to give 58 00:03:07,400 --> 00:03:10,720 Speaker 3: their views. Other students jump up and explain their theories 59 00:03:10,720 --> 00:03:13,880 Speaker 3: on the blackboard. For the next fifteen minutes, she defintely 60 00:03:13,960 --> 00:03:16,880 Speaker 3: guides the class through an in depth investigation of what 61 00:03:16,960 --> 00:03:21,480 Speaker 3: she calls shawn numbers until Sewn himself realizes that the 62 00:03:21,520 --> 00:03:24,160 Speaker 3: real meaning of odd and even is something different than 63 00:03:24,160 --> 00:03:27,160 Speaker 3: he had imagined. And now he gets it. 64 00:03:27,560 --> 00:03:30,200 Speaker 5: I've been great, Thank you for when you are in love. 65 00:03:31,120 --> 00:03:33,640 Speaker 3: I don't want to focus just on how little Sean 66 00:03:33,960 --> 00:03:36,480 Speaker 3: finally made his own way to the right answer. I'm 67 00:03:36,520 --> 00:03:38,839 Speaker 3: interested in what his teacher did to get him there. 68 00:03:39,600 --> 00:03:43,520 Speaker 3: Deborah Ball worked magic. She never told Sean the right answer, 69 00:03:44,160 --> 00:03:46,200 Speaker 3: She just led him to a place where he could 70 00:03:46,240 --> 00:03:51,120 Speaker 3: discover it for himself. My name is Malcolm Glawo. This 71 00:03:51,240 --> 00:03:54,080 Speaker 3: is season six of Smart Talks with IBM, where we 72 00:03:54,160 --> 00:03:56,720 Speaker 3: offer our listeners a glimpse behind the curtain of the 73 00:03:56,760 --> 00:04:02,000 Speaker 3: world of technology and artificial intelligence. In this season, we're 74 00:04:02,000 --> 00:04:05,200 Speaker 3: going to visit companies as varied as Laurel and Ferrari 75 00:04:05,440 --> 00:04:09,360 Speaker 3: and tell stories of how they're using artificial intelligence and 76 00:04:09,480 --> 00:04:13,800 Speaker 3: data to transform the way they do business. This episode 77 00:04:14,160 --> 00:04:16,520 Speaker 3: is about the promise of a radical new idea called 78 00:04:16,600 --> 00:04:19,919 Speaker 3: responsive teaching, the kind of teaching that took place that 79 00:04:20,080 --> 00:04:24,320 Speaker 3: day in Shawn's classroom, and whether artificial intelligence can help 80 00:04:24,400 --> 00:04:27,200 Speaker 3: us train the next generation of teachers to be as 81 00:04:27,240 --> 00:04:35,640 Speaker 3: good as Deborah Ball. Before we talk about how AI 82 00:04:35,720 --> 00:04:38,400 Speaker 3: could transform the way we train teachers, I want to 83 00:04:38,440 --> 00:04:40,920 Speaker 3: go back for a moment to the famous video of Sean. 84 00:04:41,800 --> 00:04:45,120 Speaker 3: In the video, the teacher Deborah Ball doesn't have a 85 00:04:45,200 --> 00:04:49,640 Speaker 3: predetermined plan that she's imposing on the class. She's improvising, 86 00:04:50,360 --> 00:04:53,760 Speaker 3: making up her approach as she goes along, responding to 87 00:04:53,800 --> 00:04:58,040 Speaker 3: her student's odd theory about the number six. Second, she's 88 00:04:58,080 --> 00:05:03,520 Speaker 3: taking Sean seriously. She's not dismissing his theory. She's listening 89 00:05:03,520 --> 00:05:06,839 Speaker 3: to him and trying to understand the problem from his perspective. 90 00:05:07,600 --> 00:05:11,440 Speaker 3: And Thirdly, and most importantly, she's not force feeding him 91 00:05:11,440 --> 00:05:15,359 Speaker 3: the right answer. She's being patient. She's waiting to see 92 00:05:15,400 --> 00:05:18,440 Speaker 3: if with just the right subtle hints, he can get 93 00:05:18,440 --> 00:05:23,880 Speaker 3: to the right answer on his own. Improvisation empathy patients. 94 00:05:24,720 --> 00:05:26,520 Speaker 3: That's responsive teaching. 95 00:05:26,800 --> 00:05:29,520 Speaker 7: What I think about in terms of responsiveness is more like, 96 00:05:30,640 --> 00:05:33,200 Speaker 7: I think that students need to have a sense of 97 00:05:34,200 --> 00:05:38,560 Speaker 7: agency in what happens in the classroom, and like authentic 98 00:05:38,680 --> 00:05:43,800 Speaker 7: agency where they can be legitimized as knowers. 99 00:05:44,520 --> 00:05:47,920 Speaker 3: I spoke to a physicist at Seattle Pacific University named 100 00:05:47,960 --> 00:05:52,359 Speaker 3: Amy Robertson, a longtime advocate for responsive teaching. She uses 101 00:05:52,400 --> 00:05:54,000 Speaker 3: the Sean video in her classroom. 102 00:05:54,240 --> 00:05:56,440 Speaker 7: You have to trust that kids have a way of 103 00:05:56,520 --> 00:05:59,640 Speaker 7: doing that and that, like heard, what she mostly did 104 00:05:59,720 --> 00:06:02,359 Speaker 7: was to facilitate a conversation and to say you have 105 00:06:02,400 --> 00:06:03,520 Speaker 7: to listen to them talk. 106 00:06:03,920 --> 00:06:06,440 Speaker 3: No one told him he was wrong, that's right, and 107 00:06:06,480 --> 00:06:09,560 Speaker 3: then he goes, He goes. I didn't think of it 108 00:06:09,600 --> 00:06:10,560 Speaker 3: that way again. 109 00:06:10,920 --> 00:06:13,560 Speaker 7: I thank you for bringing it up. 110 00:06:14,000 --> 00:06:17,880 Speaker 3: You've expanded my understanding. Thank you for bringing it up again. 111 00:06:17,920 --> 00:06:20,719 Speaker 3: It's like this, I love, I love. 112 00:06:21,400 --> 00:06:24,680 Speaker 7: I Responsive teaching, as I think about it, is kind 113 00:06:24,720 --> 00:06:28,480 Speaker 7: of rooted in this like Eleanor Duckworth's work around the 114 00:06:28,520 --> 00:06:31,400 Speaker 7: Having of Wonderful Ideas, where she says, like, the goal 115 00:06:31,440 --> 00:06:35,120 Speaker 7: of education is for students to have wonderful ideas and 116 00:06:35,120 --> 00:06:36,520 Speaker 7: to have a good time having them. 117 00:06:36,720 --> 00:06:40,080 Speaker 3: But I love that. I've never heard that. What a beautiful, 118 00:06:40,600 --> 00:06:44,080 Speaker 3: succinct way of summing up the purpose of education. Yes, 119 00:06:45,040 --> 00:06:48,760 Speaker 3: responsive teaching is beautiful. It's rare to find a new 120 00:06:48,800 --> 00:06:52,440 Speaker 3: teaching idea that everyone loves. This is one of those 121 00:06:52,520 --> 00:06:55,680 Speaker 3: rare ideas. Watching the Deborah Ball classroom, all I could 122 00:06:55,680 --> 00:06:58,400 Speaker 3: think was, I really really hope my daughters get to 123 00:06:58,440 --> 00:07:01,760 Speaker 3: experience a math class like that. Far too many kids 124 00:07:01,920 --> 00:07:05,040 Speaker 3: are convincing themselves at far too young an age that 125 00:07:05,120 --> 00:07:08,920 Speaker 3: math isn't for them, and responsive teaching is a way 126 00:07:09,000 --> 00:07:13,200 Speaker 3: to solve that problem. But here is the issue. It's 127 00:07:13,480 --> 00:07:17,360 Speaker 3: really really hard to teach responsive teaching. Robertson says that 128 00:07:17,440 --> 00:07:20,600 Speaker 3: teaching exists in a cultural environment where the teacher is 129 00:07:20,640 --> 00:07:23,720 Speaker 3: expected to be the source of truth that teaching is 130 00:07:23,760 --> 00:07:27,000 Speaker 3: about the immediate correction of error and not letting a 131 00:07:27,080 --> 00:07:31,760 Speaker 3: child wander down the pathway of their own misunderstanding. Responsive 132 00:07:31,800 --> 00:07:36,120 Speaker 3: teaching is deeply counterintuitive, and the only way to understand 133 00:07:36,160 --> 00:07:39,760 Speaker 3: its beauty is to do it over and over again. 134 00:07:40,680 --> 00:07:45,600 Speaker 3: Aspiring teachers need a way to practice. For as long 135 00:07:45,640 --> 00:07:48,480 Speaker 3: as there has been technology, people have turned to digital 136 00:07:48,520 --> 00:07:52,560 Speaker 3: machines to solve problems. My father was a mathematician and 137 00:07:52,600 --> 00:07:55,160 Speaker 3: I remember him coming home in the nineteen seventies with 138 00:07:55,240 --> 00:07:58,280 Speaker 3: a big stack of computer cards in his briefcase that 139 00:07:58,360 --> 00:08:01,840 Speaker 3: he used to program the main frame back the office. Today, 140 00:08:02,120 --> 00:08:06,280 Speaker 3: with the rise of artificial intelligence, the scale and complexity 141 00:08:06,520 --> 00:08:09,720 Speaker 3: of the problems technology can help us solve has jumped 142 00:08:09,760 --> 00:08:12,960 Speaker 3: by many orders of magnitude. You must have worked with 143 00:08:13,000 --> 00:08:16,800 Speaker 3: a with a million customers who are experimenting with lll ms. 144 00:08:16,840 --> 00:08:19,800 Speaker 3: Has there been one use case that you were like, WHOA, 145 00:08:19,880 --> 00:08:23,720 Speaker 3: I had no idea or just simply that's clever. I'm 146 00:08:23,800 --> 00:08:27,480 Speaker 3: speaking to Brian Bissel, who works out of IBM's Manhattan office. 147 00:08:27,880 --> 00:08:31,080 Speaker 3: He helps IBM customers discover how best to get AI 148 00:08:31,200 --> 00:08:32,080 Speaker 3: to work for them. 149 00:08:32,800 --> 00:08:34,480 Speaker 8: There is one, but I don't think I can talk 150 00:08:34,480 --> 00:08:34,800 Speaker 8: about it. 151 00:08:34,880 --> 00:08:38,240 Speaker 3: Unfortunately, Wait, wait, you can't tease me like that, can 152 00:08:38,320 --> 00:08:42,320 Speaker 3: you wait? Disguise disguise it for me, Just give me 153 00:08:42,360 --> 00:08:43,200 Speaker 3: a general. 154 00:08:43,480 --> 00:08:45,280 Speaker 8: It was about the ability to pull certain types of 155 00:08:45,280 --> 00:08:49,000 Speaker 8: information out of documents that you you wouldn't think you 156 00:08:49,000 --> 00:08:52,320 Speaker 8: would be able to get the model to do, and 157 00:08:52,480 --> 00:08:55,440 Speaker 8: be able to do that at a very large scale. 158 00:08:55,800 --> 00:08:58,440 Speaker 3: Bissile's point was that we are well past the stage 159 00:08:58,480 --> 00:09:01,760 Speaker 3: where anyone wonders whether a I can be useful. The 160 00:09:01,800 --> 00:09:04,920 Speaker 3: real question now is what problems do we want to 161 00:09:05,000 --> 00:09:07,720 Speaker 3: use it to solve? Where it can make the biggest difference, 162 00:09:08,440 --> 00:09:11,800 Speaker 3: And Basil saw lots of opportunities in education. 163 00:09:12,720 --> 00:09:15,800 Speaker 8: I have two kids, one in middle school and one 164 00:09:15,880 --> 00:09:19,160 Speaker 8: who just graduated high school, and I'm well aware of 165 00:09:19,240 --> 00:09:22,840 Speaker 8: students using things like chat GPT to do their homework. 166 00:09:23,640 --> 00:09:29,080 Speaker 8: And it's very easy to take tools like that and 167 00:09:29,120 --> 00:09:32,800 Speaker 8: even IBM's own large language models and just take a problem, 168 00:09:33,160 --> 00:09:36,880 Speaker 8: a piece of homework, something you want written, and drop 169 00:09:36,920 --> 00:09:38,800 Speaker 8: it into that and have it generate the answer for 170 00:09:38,880 --> 00:09:43,199 Speaker 8: you and the student. The user in that case hasn't 171 00:09:43,240 --> 00:09:46,240 Speaker 8: done any work, they haven't put any real thought into it. 172 00:09:46,280 --> 00:09:50,240 Speaker 3: To Basil, that's the wrong use of AI that's technology 173 00:09:50,480 --> 00:09:54,280 Speaker 3: making is dumber. What we really want is technology that 174 00:09:54,360 --> 00:09:57,280 Speaker 3: makes us smarter. Basil explain to me that there are 175 00:09:57,320 --> 00:10:00,840 Speaker 3: now two big tools being used for AI productivity, AI 176 00:10:01,000 --> 00:10:06,320 Speaker 3: agents and AI assistants. Let's start with AI agents. AI 177 00:10:06,400 --> 00:10:10,400 Speaker 3: agents can reason plan and collaborate with other AI tools 178 00:10:10,640 --> 00:10:14,640 Speaker 3: to autonomously perform tasks for a user. This will gave 179 00:10:14,679 --> 00:10:17,200 Speaker 3: me an example of how college freshmen might use an 180 00:10:17,200 --> 00:10:17,800 Speaker 3: AI agent. 181 00:10:18,200 --> 00:10:20,720 Speaker 8: As a new student, you may not know how do 182 00:10:20,760 --> 00:10:22,839 Speaker 8: I do with my health and wellness issue? So many 183 00:10:22,880 --> 00:10:25,440 Speaker 8: credits are going to get for this given class. You 184 00:10:25,480 --> 00:10:28,839 Speaker 8: could talk to someone and find out some of that, 185 00:10:29,840 --> 00:10:31,440 Speaker 8: but maybe it's a little bit sensitive and you don't 186 00:10:31,440 --> 00:10:31,959 Speaker 8: want to do that. 187 00:10:32,480 --> 00:10:35,640 Speaker 3: Bisill told me you could build an AI agent, a 188 00:10:35,720 --> 00:10:39,160 Speaker 3: resource for new students that helps them navigate a new campus, 189 00:10:39,480 --> 00:10:42,720 Speaker 3: register for classes, access the services they need, and even 190 00:10:42,720 --> 00:10:46,559 Speaker 3: schedule appointments on their behalf, which in turn buys them 191 00:10:46,559 --> 00:10:48,880 Speaker 3: more time to focus on their actual school work. 192 00:10:49,200 --> 00:10:54,120 Speaker 8: We can see patterns of how agents and assistants can 193 00:10:54,200 --> 00:10:58,599 Speaker 8: help employees and customers and end users be more productive. 194 00:10:58,800 --> 00:11:03,040 Speaker 8: Automate workflows are not doing certain types of repetitive work 195 00:11:03,520 --> 00:11:07,240 Speaker 8: over and over again. And streamlining their lives and making 196 00:11:07,360 --> 00:11:10,880 Speaker 8: data more accessible to them twenty four hours a day. 197 00:11:11,240 --> 00:11:14,960 Speaker 3: But Bissil says you can also use AI assistance in 198 00:11:15,040 --> 00:11:19,160 Speaker 3: the education space. AI assistants are reactive as opposed to 199 00:11:19,200 --> 00:11:23,559 Speaker 3: AI agents, which are proactive. AI assistants only performed tasks 200 00:11:23,600 --> 00:11:28,520 Speaker 3: at your request. They're programmed to answer your questions, and 201 00:11:28,559 --> 00:11:31,640 Speaker 3: as it turns out, AI assistants are now being used 202 00:11:31,960 --> 00:11:35,760 Speaker 3: to further the responsive teaching revolution, which is why I 203 00:11:35,800 --> 00:11:39,280 Speaker 3: found myself on a beautiful Georgia spring day not long ago, 204 00:11:39,440 --> 00:11:43,040 Speaker 3: on the campus of Kansas State University, sitting in a 205 00:11:43,040 --> 00:11:47,200 Speaker 3: classroom with two researchers, one of them Professor Dabe Lee. 206 00:11:47,440 --> 00:11:49,920 Speaker 3: Let's go into the journey of building this thing. You 207 00:11:50,480 --> 00:11:52,040 Speaker 3: started w by taking a course. 208 00:11:52,040 --> 00:11:54,960 Speaker 9: What was the course you took, Yeah, so it was 209 00:11:55,520 --> 00:12:01,959 Speaker 9: offered by Coursera. It was designed by IBM AI Foundation 210 00:12:02,240 --> 00:12:03,040 Speaker 9: for everyone. 211 00:12:06,320 --> 00:12:09,000 Speaker 3: In her AI Foundation's course, Lee learned how to build 212 00:12:09,000 --> 00:12:12,719 Speaker 3: an AI assistant using IBM Watson X. That course took 213 00:12:12,760 --> 00:12:14,000 Speaker 3: how long to take. 214 00:12:14,520 --> 00:12:17,199 Speaker 9: It was not to know it was like fourteen weeks. 215 00:12:18,240 --> 00:12:21,640 Speaker 3: Lee's idea was to train an AI assistant on classroom 216 00:12:21,720 --> 00:12:25,120 Speaker 3: data to play the role of sean A digital persona 217 00:12:25,360 --> 00:12:27,839 Speaker 3: of a nine year old who likes math but doesn't 218 00:12:27,840 --> 00:12:31,840 Speaker 3: always understand math, and that AI assistant she thought could 219 00:12:31,880 --> 00:12:35,000 Speaker 3: be used to train preservice teachers or teachers in training 220 00:12:35,280 --> 00:12:37,880 Speaker 3: who are preparing to enter one of the most challenging 221 00:12:37,880 --> 00:12:39,640 Speaker 3: professions in the modern world. 222 00:12:40,600 --> 00:12:43,600 Speaker 9: So when you think about the teacher education and a 223 00:12:43,800 --> 00:12:49,439 Speaker 9: major challenge that teacher education phase is that we need 224 00:12:49,679 --> 00:12:54,400 Speaker 9: children to practice with. We need instructors who will give 225 00:12:54,520 --> 00:12:59,560 Speaker 9: the instruction on the pedagogical skills. So when you look 226 00:12:59,600 --> 00:13:03,839 Speaker 9: at the education program, we have coursework in field experience, 227 00:13:04,360 --> 00:13:08,720 Speaker 9: and in those two areas there is something missing all 228 00:13:08,800 --> 00:13:09,240 Speaker 9: the time. 229 00:13:10,400 --> 00:13:13,760 Speaker 3: Li says that pre service teachers often lack access to 230 00:13:13,840 --> 00:13:17,360 Speaker 3: both students and experienced teachers during their education. 231 00:13:18,000 --> 00:13:22,400 Speaker 9: So what we try to resolve is that we have 232 00:13:22,520 --> 00:13:25,760 Speaker 9: this virtual student for pre service teacher to work with 233 00:13:26,559 --> 00:13:30,800 Speaker 9: so that they can practice their responsive teaching skills. 234 00:13:31,000 --> 00:13:35,080 Speaker 3: The first AI assistant Lee created is g Wu gi Wu, 235 00:13:35,240 --> 00:13:39,160 Speaker 3: emulates the persona of a nine year old third grade girl. Then, 236 00:13:39,200 --> 00:13:41,880 Speaker 3: with the help of one of her collaborators, a researcher 237 00:13:41,920 --> 00:13:46,199 Speaker 3: at Canazon named Sean English, she created two more AI assistants, 238 00:13:46,600 --> 00:13:51,679 Speaker 3: Gabriel and Noah, each of which have their own distinctive characteristics. 239 00:13:51,920 --> 00:13:55,160 Speaker 3: So how are gabriel and Noah different from. 240 00:13:55,080 --> 00:14:01,120 Speaker 9: G Wu Gabrielle? My first one is very short answered. 241 00:14:01,360 --> 00:14:04,240 Speaker 9: If you ask an open ended question, he will answer 242 00:14:04,280 --> 00:14:08,679 Speaker 9: it in a close way. So I use that characteristic. 243 00:14:08,760 --> 00:14:13,800 Speaker 9: And that's the problem that most teachers actually base. They're 244 00:14:13,840 --> 00:14:17,040 Speaker 9: asked children who are shay, who are reserved, and who 245 00:14:17,040 --> 00:14:21,680 Speaker 9: would not share much of their thoughts. So we wanted 246 00:14:21,760 --> 00:14:25,520 Speaker 9: that characteristic in some characters, and we use Gabrielle to 247 00:14:25,840 --> 00:14:27,000 Speaker 9: have that characteristic. 248 00:14:28,480 --> 00:14:30,960 Speaker 3: And Noah. What'snawah's personality? 249 00:14:32,320 --> 00:14:36,040 Speaker 6: How do he playful? Cheery, bright and energetic? 250 00:14:36,840 --> 00:14:39,600 Speaker 3: That's Sean English professor, Lee's fellow researcher. 251 00:14:40,120 --> 00:14:46,640 Speaker 9: And Jewu is articulated and kind of smart, but he 252 00:14:46,840 --> 00:14:48,840 Speaker 9: she has her own way of thinking. 253 00:14:49,240 --> 00:14:51,160 Speaker 3: I would end up spending a lot of time with 254 00:14:51,280 --> 00:14:54,640 Speaker 3: jie Wu. She's something of a character. I asked Sean 255 00:14:54,680 --> 00:14:57,920 Speaker 3: about the process of creating these AI assistants. What does 256 00:14:58,040 --> 00:15:03,560 Speaker 3: building the content side of the AI assistant entail? Sean? 257 00:15:03,560 --> 00:15:03,600 Speaker 9: What? 258 00:15:04,000 --> 00:15:06,480 Speaker 6: It sets up a series of actions, effectively, which are 259 00:15:07,160 --> 00:15:09,680 Speaker 6: response cases. You can kind of think of them as 260 00:15:09,960 --> 00:15:12,760 Speaker 6: you have a series of questions that you tie to 261 00:15:13,920 --> 00:15:17,720 Speaker 6: an intent, and then that intent has reactions from the bot, 262 00:15:18,040 --> 00:15:21,160 Speaker 6: and so effectively, if we were looking to say make 263 00:15:21,200 --> 00:15:23,520 Speaker 6: a hello action, we would have all the different ways 264 00:15:23,520 --> 00:15:25,880 Speaker 6: that people could say Hello, Hello, what's up, how you doing, 265 00:15:25,880 --> 00:15:26,880 Speaker 6: and all that kind of stuff. 266 00:15:27,240 --> 00:15:31,120 Speaker 3: Sean says, the longer the list of potential responses, the better, 267 00:15:31,600 --> 00:15:35,520 Speaker 3: But AI's responses don't just follow the list. The AI 268 00:15:35,560 --> 00:15:39,120 Speaker 3: assistant uses those suggested responses to come up with a 269 00:15:39,240 --> 00:15:43,120 Speaker 3: universe of other responses, and in that process sometimes it 270 00:15:43,160 --> 00:15:45,480 Speaker 3: comes up with things that just don't make sense. 271 00:15:45,560 --> 00:15:49,040 Speaker 6: And from a technological standpoint, while AI is a fantastic tool, 272 00:15:49,160 --> 00:15:51,600 Speaker 6: AI can hallucinate, which I mean, just give things that 273 00:15:51,600 --> 00:15:54,680 Speaker 6: it's just straight up made up. There's a famous example 274 00:15:54,720 --> 00:15:56,640 Speaker 6: of this called the three rs is where you ask 275 00:15:56,680 --> 00:15:59,640 Speaker 6: a popular large language model how many RS are in strawberry, 276 00:16:00,000 --> 00:16:01,960 Speaker 6: and it gives you the wrong answer, and he repeats 277 00:16:02,000 --> 00:16:04,800 Speaker 6: that result repetitively. You always want to have a human 278 00:16:05,320 --> 00:16:07,320 Speaker 6: interacting with the system to be able to go, hey, 279 00:16:07,920 --> 00:16:10,080 Speaker 6: that's a little crazy. I don't think that's exactly what 280 00:16:10,080 --> 00:16:10,920 Speaker 6: we're going for here. 281 00:16:11,560 --> 00:16:13,560 Speaker 3: That's why it's good to have someone like Sean English 282 00:16:13,560 --> 00:16:15,640 Speaker 3: around to step in and get the model back on track, 283 00:16:16,040 --> 00:16:19,200 Speaker 3: and over time, when the model has enough training, it's 284 00:16:19,280 --> 00:16:25,640 Speaker 3: ready for the teachers in training. One of the rollouts 285 00:16:25,640 --> 00:16:28,720 Speaker 3: of Jiwu, Gabriel, and Noah was with the teacher training 286 00:16:28,760 --> 00:16:30,720 Speaker 3: program at the University of Missouri. 287 00:16:31,040 --> 00:16:33,400 Speaker 10: I was just kind of excited to see what the 288 00:16:33,480 --> 00:16:35,880 Speaker 10: program was and what it was going to be doing. 289 00:16:36,360 --> 00:16:39,400 Speaker 3: This is Logan Hovis, a junior at Missouri on the 290 00:16:39,440 --> 00:16:41,760 Speaker 3: path to becoming an elementary school teacher. 291 00:16:42,160 --> 00:16:44,240 Speaker 11: Obviously a little skeptical when he said it was so 292 00:16:44,520 --> 00:16:47,640 Speaker 11: to you know, be like talking to a student. You're like, 293 00:16:47,640 --> 00:16:50,040 Speaker 11: there's no way this AI thing is going to totally 294 00:16:50,080 --> 00:16:52,760 Speaker 11: sound like a second grader or a third grader, Like 295 00:16:52,800 --> 00:16:55,360 Speaker 11: it's going to sound like an adult, or it's going 296 00:16:55,400 --> 00:16:57,240 Speaker 11: to sound like a robot that knows all the answers. 297 00:16:57,760 --> 00:17:00,480 Speaker 10: And it really didn't. It really was like talking to 298 00:17:00,520 --> 00:17:00,880 Speaker 10: a child. 299 00:17:00,920 --> 00:17:03,480 Speaker 11: It was very very well developed in the way that 300 00:17:03,520 --> 00:17:05,600 Speaker 11: you really sit there and you feel like you're talking 301 00:17:05,600 --> 00:17:06,120 Speaker 11: to a kid. 302 00:17:06,720 --> 00:17:09,920 Speaker 3: Her point wasn't that Jiwu and her fellow avatars were 303 00:17:09,960 --> 00:17:13,720 Speaker 3: equivalent to real kids. Of course not, but for someone 304 00:17:13,840 --> 00:17:16,760 Speaker 3: starting out, someone who was already nervous about being plunged 305 00:17:16,760 --> 00:17:19,480 Speaker 3: into a classroom of nine year olds, Jeewu was like 306 00:17:19,520 --> 00:17:21,400 Speaker 3: a warm up before a baseball game. 307 00:17:21,800 --> 00:17:23,919 Speaker 11: What I can think of is like, you know how 308 00:17:24,320 --> 00:17:26,600 Speaker 11: when you're at batting practice for baseball or softball, you 309 00:17:26,640 --> 00:17:29,560 Speaker 11: have those automatic pitchers that throw them because you're working 310 00:17:29,640 --> 00:17:32,280 Speaker 11: on your skill as the hitter. What can I do differently? 311 00:17:32,280 --> 00:17:35,320 Speaker 11: What am I doing wrong? But that doesn't replace the 312 00:17:35,359 --> 00:17:37,080 Speaker 11: game and what you do in a game. But this 313 00:17:37,200 --> 00:17:40,000 Speaker 11: is you getting to practice your own skills to be 314 00:17:40,000 --> 00:17:41,480 Speaker 11: better when you go in a game. And I think 315 00:17:41,520 --> 00:17:44,600 Speaker 11: that's kind of what the AI software feels like for us. 316 00:17:46,440 --> 00:17:49,119 Speaker 3: In batting practice, the pitches don't come as hard and 317 00:17:49,160 --> 00:17:51,679 Speaker 3: fast as the pitch is in a real game, but 318 00:17:51,720 --> 00:17:53,560 Speaker 3: you get to stand at the plate and the pitcher 319 00:17:53,640 --> 00:17:57,320 Speaker 3: throws you dozens of balls over and over again in 320 00:17:57,359 --> 00:17:59,959 Speaker 3: a concentrated block that allows you to work on your 321 00:18:00,160 --> 00:18:02,760 Speaker 3: swing closely and carefully. 322 00:18:03,320 --> 00:18:05,760 Speaker 10: There's a lot less stimulus going on around because the 323 00:18:05,800 --> 00:18:08,560 Speaker 10: classroom is very very busy. It's wonderful, it's beautiful, but 324 00:18:08,600 --> 00:18:11,080 Speaker 10: it's very very busy, so sometimes it's hard to keep 325 00:18:11,600 --> 00:18:14,880 Speaker 10: you know, that focus in on the tasks that they're 326 00:18:14,920 --> 00:18:17,479 Speaker 10: doing at hand, and also in the teacher setting, you're 327 00:18:17,520 --> 00:18:20,160 Speaker 10: also kind of always looking around making sure that other 328 00:18:20,200 --> 00:18:22,240 Speaker 10: students are doing what they're supposed to be doing, but 329 00:18:22,320 --> 00:18:24,440 Speaker 10: also like if they need any help, if everything's going 330 00:18:24,480 --> 00:18:29,760 Speaker 10: okay in the classroom. So being on the Jiwu chat, 331 00:18:30,240 --> 00:18:32,119 Speaker 10: it was just nice that you didn't have to do 332 00:18:32,280 --> 00:18:34,600 Speaker 10: any of the extra work to keep the focus on there, 333 00:18:35,040 --> 00:18:38,600 Speaker 10: and it also felt you did have to feel the 334 00:18:38,640 --> 00:18:42,240 Speaker 10: student's nervousness of being one on one with you, and 335 00:18:42,440 --> 00:18:44,800 Speaker 10: also as a teacher, it was a lot less pressure 336 00:18:44,840 --> 00:18:47,520 Speaker 10: too because I was like, Okay, I'm taking this series. 337 00:18:47,560 --> 00:18:49,840 Speaker 10: This is a student I'm questioning, but. 338 00:18:49,800 --> 00:18:52,399 Speaker 11: I also know I'm probably not going to hurt someone's 339 00:18:52,400 --> 00:18:55,040 Speaker 11: feelings right now, and that's terrifying to think I'm going 340 00:18:55,080 --> 00:18:58,640 Speaker 11: to ask the wrong question and upset the child because 341 00:18:58,680 --> 00:18:59,479 Speaker 11: I've done that. 342 00:19:00,440 --> 00:19:02,879 Speaker 3: We think of the typical use of AI as a 343 00:19:02,920 --> 00:19:05,680 Speaker 3: tool for speeding things up. That's what we always hear 344 00:19:05,960 --> 00:19:08,879 Speaker 3: that the introduction of AI to problem X gave an 345 00:19:08,880 --> 00:19:13,159 Speaker 3: answer in minutes when solving problem X used to take weeks. 346 00:19:13,720 --> 00:19:17,640 Speaker 3: But we shouldn't forget another use that it allows us 347 00:19:17,960 --> 00:19:21,360 Speaker 3: to slow things down. Hoves, if she wanted to, could 348 00:19:21,359 --> 00:19:24,560 Speaker 3: spend a whole weekend practicing with ji Wu. A real 349 00:19:24,640 --> 00:19:27,320 Speaker 3: nine year old will get frustrated on board with the 350 00:19:27,320 --> 00:19:31,320 Speaker 3: fumbling novice after ten minutes, but gi Wu ji Wu 351 00:19:31,480 --> 00:19:34,320 Speaker 3: will happily answer questions for as long as it takes 352 00:19:34,320 --> 00:19:36,920 Speaker 3: for the people who want to learn to be responsive 353 00:19:37,520 --> 00:19:41,760 Speaker 3: to learn how to be responsive. At the end of 354 00:19:41,800 --> 00:19:45,280 Speaker 3: my time at Kennesas State, Sean and Dabe led me 355 00:19:45,359 --> 00:19:48,320 Speaker 3: to a small table where Dabe had set up her laptop. 356 00:19:48,920 --> 00:19:51,240 Speaker 3: In the corner of the screen was a chat box 357 00:19:51,520 --> 00:19:53,960 Speaker 3: of the sort we've all seen and used a thousand times. 358 00:19:54,720 --> 00:19:58,240 Speaker 3: Ji Wu began. She had been given a math problem. 359 00:19:58,800 --> 00:20:04,400 Speaker 4: Rutle, who are of three fourth cup of a flower 360 00:20:04,720 --> 00:20:09,480 Speaker 4: to the ball thanks to the added another three six 361 00:20:09,680 --> 00:20:14,480 Speaker 4: is cup. It's a total amount of flower the use 362 00:20:15,040 --> 00:20:20,240 Speaker 4: greater or dan or a less than one cup? How 363 00:20:20,320 --> 00:20:21,560 Speaker 4: much flower. 364 00:20:21,280 --> 00:20:22,120 Speaker 10: They can use. 365 00:20:22,520 --> 00:20:25,600 Speaker 3: That's a simulation of Giewu speaking. We pause it for 366 00:20:25,600 --> 00:20:31,080 Speaker 3: a second. So Jewu is trying to solve this problem. 367 00:20:31,160 --> 00:20:33,760 Speaker 3: And the first thing she does is she draws a 368 00:20:33,840 --> 00:20:37,440 Speaker 3: rectangle on the screen. This is a common tactic of 369 00:20:37,480 --> 00:20:41,600 Speaker 3: nine year olds try to visualize the fractions. And she 370 00:20:41,680 --> 00:20:47,040 Speaker 3: divides it into four pieces. And now she's gonna color 371 00:20:47,119 --> 00:20:49,760 Speaker 3: in three of the four pieces. Yes, so she's representing 372 00:20:49,760 --> 00:20:52,760 Speaker 3: this is quite good. She's representing three quarters on the screen. 373 00:20:54,440 --> 00:20:58,080 Speaker 4: This is three sixes. 374 00:21:00,080 --> 00:21:05,520 Speaker 3: So now Jiwu does another rectangle with six boxes and 375 00:21:05,600 --> 00:21:06,840 Speaker 3: colors in three of them. 376 00:21:07,160 --> 00:21:13,960 Speaker 4: Okay, together makes sikes going off. 377 00:21:15,560 --> 00:21:19,520 Speaker 3: So then she counts up all the colored boxes and 378 00:21:19,560 --> 00:21:22,879 Speaker 3: that's her numerator, and counts up the total number of 379 00:21:22,880 --> 00:21:26,520 Speaker 3: boxes and that's her denominator. Ji Wu had counted the 380 00:21:26,560 --> 00:21:30,160 Speaker 3: colored boxes and landed on an answer. When you add 381 00:21:30,400 --> 00:21:33,080 Speaker 3: three quarters of a cup and three sixths of a cup, 382 00:21:33,680 --> 00:21:36,800 Speaker 3: you get six tenths of a cup. So, according to 383 00:21:36,880 --> 00:21:39,800 Speaker 3: ji Wu, Martin has less than one cup. And she 384 00:21:39,920 --> 00:21:41,040 Speaker 3: thinks she solved the problem. 385 00:21:41,200 --> 00:21:43,879 Speaker 9: Yes, okay, so it's less than one cup. 386 00:21:44,200 --> 00:21:47,480 Speaker 3: Yeah, so she says it's less than one cup. Now, 387 00:21:47,560 --> 00:21:50,480 Speaker 3: oh my god, this is hard. So the question is 388 00:21:50,520 --> 00:21:53,880 Speaker 3: what do I, as a teacher say to Jiwu. We 389 00:21:53,880 --> 00:21:57,000 Speaker 3: were off. The rules were simple. I couldn't give ji 390 00:21:57,040 --> 00:21:59,560 Speaker 3: Wu the answer or explain to her what she was 391 00:21:59,560 --> 00:22:02,520 Speaker 3: doing wrong. I had to be Deborah Ball. I had 392 00:22:02,520 --> 00:22:05,879 Speaker 3: to help her find the way herself. The chat box 393 00:22:06,000 --> 00:22:08,040 Speaker 3: in the corner of the screen was waiting for my 394 00:22:08,080 --> 00:22:11,359 Speaker 3: first question. I thought for a moment and started typing, 395 00:22:11,680 --> 00:22:14,720 Speaker 3: do you think the boxes in the red rectangle are 396 00:22:14,760 --> 00:22:17,800 Speaker 3: the same size as the boxes in the blue rectangle. 397 00:22:18,640 --> 00:22:21,040 Speaker 3: Then I turned to Sean and dabey, is that a 398 00:22:21,040 --> 00:22:21,560 Speaker 3: good question? 399 00:22:21,800 --> 00:22:23,280 Speaker 2: Yeah, serious thing. 400 00:22:24,280 --> 00:22:25,719 Speaker 9: Yeah, that's a good question. 401 00:22:26,240 --> 00:22:30,359 Speaker 3: Jewu doesn't mess around. She answers immediately. So Ju says, 402 00:22:30,359 --> 00:22:32,560 Speaker 3: the blue and red pieces are not the same sizes. 403 00:22:33,359 --> 00:22:37,639 Speaker 9: Oh so you understand now, gu knows that side differences. 404 00:22:39,600 --> 00:22:40,640 Speaker 3: She's pretty smart here. 405 00:22:40,800 --> 00:22:41,480 Speaker 9: Yeah. 406 00:22:41,600 --> 00:22:44,320 Speaker 3: Then I asked, if they are not the same size, 407 00:22:44,640 --> 00:22:47,960 Speaker 3: do you think you can add them together? Jiwu answered 408 00:22:48,000 --> 00:22:51,880 Speaker 3: right away. Jiwu says, I have learned that I could 409 00:22:51,920 --> 00:22:54,920 Speaker 3: add any numbers in grade two. So three p three 410 00:22:55,000 --> 00:22:56,720 Speaker 3: is six and four to six is ten. 411 00:22:57,040 --> 00:23:01,480 Speaker 9: Yeah, so she is using the knowledge of edge intiquers 412 00:23:01,520 --> 00:23:03,280 Speaker 9: into adding fractions. 413 00:23:03,840 --> 00:23:07,760 Speaker 3: Now I'm stumped. So now I have to somehow lead 414 00:23:07,880 --> 00:23:10,439 Speaker 3: her to figure out a way to get her to 415 00:23:10,520 --> 00:23:15,200 Speaker 3: understand that we're dealing with a different kind of problem, 416 00:23:15,400 --> 00:23:18,640 Speaker 3: a harder problem. Amy Robertson had told me that learning 417 00:23:18,640 --> 00:23:22,399 Speaker 3: how to do responsive teaching properly was really hard, and 418 00:23:22,440 --> 00:23:25,399 Speaker 3: now I understood why. I had to put my mind 419 00:23:25,760 --> 00:23:27,840 Speaker 3: inside the mind of a nine year old. I had 420 00:23:27,880 --> 00:23:31,399 Speaker 3: to internalize her knowledge base and assumptions and keep in mind, 421 00:23:31,720 --> 00:23:35,199 Speaker 3: I haven't been nine for a very long time. I 422 00:23:35,280 --> 00:23:38,480 Speaker 3: honestly had no idea what to say next. I thought 423 00:23:38,480 --> 00:23:41,040 Speaker 3: for a moment, I asked what I quickly realized was 424 00:23:41,040 --> 00:23:44,800 Speaker 3: a hopelessly convoluted question. Dobby and Sean had built a 425 00:23:44,800 --> 00:23:49,080 Speaker 3: mentor into the system, an experienced, responsive teacher who supervises 426 00:23:49,080 --> 00:23:52,280 Speaker 3: the session and offers advice. My mentor noticed that I 427 00:23:52,320 --> 00:23:58,800 Speaker 3: was struggling, told me to simplify my question. Grader Dobby 428 00:23:58,880 --> 00:24:01,359 Speaker 3: was trying to help me too, She suggested, why not 429 00:24:01,520 --> 00:24:05,000 Speaker 3: just ask ji Wu if three quarters is bigger or 430 00:24:05,040 --> 00:24:06,200 Speaker 3: smaller than one half? 431 00:24:06,600 --> 00:24:10,040 Speaker 9: So we are trying to help her to think about 432 00:24:10,119 --> 00:24:12,720 Speaker 9: faction in a more conceptual way. 433 00:24:12,760 --> 00:24:17,720 Speaker 3: This time, Jiwu understood. She wrote back, three quarters is 434 00:24:17,880 --> 00:24:21,520 Speaker 3: larger than one half? I wrote back, is three sixths 435 00:24:21,520 --> 00:24:25,879 Speaker 3: of a cup bigger or smaller than one half? Jewu said, 436 00:24:26,520 --> 00:24:27,320 Speaker 3: I'm confused. 437 00:24:27,680 --> 00:24:29,880 Speaker 2: Oh no, I've confused, gi Wu. 438 00:24:30,040 --> 00:24:34,720 Speaker 9: But it's good she's understanding. She's realizing her misconception. So 439 00:24:34,920 --> 00:24:35,920 Speaker 9: she's getting confused. 440 00:24:35,920 --> 00:24:38,240 Speaker 3: She says, I'm confused. Three quarters is pretty close to 441 00:24:38,280 --> 00:24:40,640 Speaker 3: one and adding three six would make it go over one. 442 00:24:42,200 --> 00:24:43,640 Speaker 2: Oh, so she's got the answer. 443 00:24:43,840 --> 00:24:46,960 Speaker 3: Yeah, But then she says, but there are six pieces 444 00:24:46,960 --> 00:24:48,520 Speaker 3: out of ten, which is less than one, so I 445 00:24:48,520 --> 00:24:49,120 Speaker 3: don't get it. 446 00:24:49,640 --> 00:24:53,280 Speaker 9: So she's the point that, oh this, I have something 447 00:24:53,400 --> 00:24:55,359 Speaker 9: wrong here. That's a good sign. 448 00:24:55,840 --> 00:24:56,560 Speaker 3: She's getting there. 449 00:24:56,600 --> 00:24:57,960 Speaker 9: Yeah, she's getting there, but. 450 00:24:57,880 --> 00:25:01,080 Speaker 3: I still have to get her. She has to get 451 00:25:01,240 --> 00:25:02,600 Speaker 3: the six pieces out of ten. 452 00:25:02,560 --> 00:25:03,160 Speaker 2: Out of her head. 453 00:25:03,359 --> 00:25:06,920 Speaker 3: Yeah, I have no idea how to do that? What 454 00:25:06,320 --> 00:25:10,480 Speaker 3: he and she thinks she's confused when she has Actually 455 00:25:11,600 --> 00:25:14,439 Speaker 3: she's figured out the answer. Yeah, she did, So we 456 00:25:14,480 --> 00:25:18,120 Speaker 3: have advance. Even in my stumbling and bumbling, we've made 457 00:25:18,200 --> 00:25:18,960 Speaker 3: some progress, and. 458 00:25:19,119 --> 00:25:19,960 Speaker 6: Very notable progress. 459 00:25:26,040 --> 00:25:28,520 Speaker 3: My conversation with jie Wu went on for some time, 460 00:25:28,800 --> 00:25:32,120 Speaker 3: and eventually I got there. Jie Wu found her way 461 00:25:32,160 --> 00:25:35,360 Speaker 3: to the right answer. She said, I have more than 462 00:25:35,359 --> 00:25:38,320 Speaker 3: one cup of flower. The mentor chimed in. I got 463 00:25:38,320 --> 00:25:40,840 Speaker 3: a little emoji that made me feel good, And when 464 00:25:40,840 --> 00:25:44,000 Speaker 3: it was over, I realized two things. The first was 465 00:25:44,240 --> 00:25:47,879 Speaker 3: I needed more batting practice, much more, and that batting 466 00:25:47,880 --> 00:25:51,440 Speaker 3: practice was really really easy to do, because someone has 467 00:25:51,440 --> 00:25:53,840 Speaker 3: gone to the trouble of building me my very own 468 00:25:53,880 --> 00:25:57,040 Speaker 3: baseball diamond and given me a pitcher who had thrown 469 00:25:57,080 --> 00:26:01,320 Speaker 3: me baseballs all day long. The second thought was that 470 00:26:01,359 --> 00:26:05,200 Speaker 3: I've been thinking about AI all wrong. I have interpreted 471 00:26:05,240 --> 00:26:07,000 Speaker 3: a lot of the talk about the promise of AI 472 00:26:07,320 --> 00:26:10,920 Speaker 3: to be about replacing human expertise. I had actually thought 473 00:26:11,040 --> 00:26:13,600 Speaker 3: when I first heard about Dabe's project that that's what 474 00:26:13,760 --> 00:26:17,080 Speaker 3: Dabe and Sean were doing, creating an AI to teach 475 00:26:17,080 --> 00:26:20,360 Speaker 3: students by passing the teacher altogether. But if you did 476 00:26:20,359 --> 00:26:23,320 Speaker 3: it that way, you had missed the magic of the classroom. 477 00:26:23,640 --> 00:26:27,879 Speaker 3: Remember Eleanor Duckworth's quote, the goal of education is for 478 00:26:27,960 --> 00:26:30,920 Speaker 3: students to have wonderful ideas and have a good time 479 00:26:31,000 --> 00:26:34,399 Speaker 3: having them. I think we often focus on the first 480 00:26:34,440 --> 00:26:38,520 Speaker 3: part of that formulation, the wonderful ideas, but neglect the second, 481 00:26:39,080 --> 00:26:44,040 Speaker 3: the good time having them. Real learning is born in pleasure, 482 00:26:44,640 --> 00:26:48,560 Speaker 3: in community, in playful discussion, in a group of kids 483 00:26:48,600 --> 00:26:51,440 Speaker 3: coming together to solve a problem, and all of that 484 00:26:51,480 --> 00:26:56,000 Speaker 3: magic only comes from human interaction from a teacher who 485 00:26:56,040 --> 00:26:58,760 Speaker 3: is skilled enough to inspire a class of nine year olds. 486 00:26:59,320 --> 00:27:03,280 Speaker 3: We don't want AI assistants to replace the teacher. We 487 00:27:03,320 --> 00:27:08,040 Speaker 3: want AI assistants to help teachers turn themselves into even 488 00:27:08,119 --> 00:27:25,960 Speaker 3: better teachers. Smart Talks with IBM is produced by Matt Romano, 489 00:27:26,359 --> 00:27:30,720 Speaker 3: Amy Gaines McQuaid Lucy Sullivan and Jake Harper. We're edited 490 00:27:30,760 --> 00:27:34,880 Speaker 3: by Lacy Roberts, Engineering by Nina Bird Lawrence, mastering by 491 00:27:34,920 --> 00:27:39,600 Speaker 3: Sarah Brugerer. Music by Gramoscope. Special thanks to Tatiana Lieberman 492 00:27:39,800 --> 00:27:43,240 Speaker 3: and Cassidy Meyer. Smart Talks with IBM is a production 493 00:27:43,320 --> 00:27:47,840 Speaker 3: of Pushkin Industries and Ruby Studio at iHeartMedia. To find 494 00:27:47,840 --> 00:27:52,600 Speaker 3: more Pushkin podcasts, listen on the iHeartRadio app, Apple Podcasts, 495 00:27:52,680 --> 00:27:56,359 Speaker 3: or wherever you get your podcasts. I'm Malcolm Gabo. This 496 00:27:56,520 --> 00:28:00,280 Speaker 3: is a paid advertisement from IBM. The conversations on this 497 00:28:00,320 --> 00:28:05,879 Speaker 3: podcast don't necessarily represent IBM's positions, strategies, or opinions.