WEBVTT - Smart Talks with IBM: How AI Assistants Can Transform Education 0:00:00.160 --> 0:00:02.960 Hey everyone, it's Robert and Joe here. Today we've got 0:00:02.960 --> 0:00:05.240 something a little bit different to share with you. It's 0:00:05.280 --> 0:00:09.400 a new season of the Smart Talks with IBM podcast series. 0:00:09.720 --> 0:00:13.520 This season on Smart Talks with IBM, Malcolm Gladwell is back, 0:00:13.520 --> 0:00:15.760 and this time he's taking the show on the road. 0:00:15.880 --> 0:00:19.160 Malcolm is stepping outside the studio to explore how IBM 0:00:19.239 --> 0:00:22.880 clients are using artificial intelligence to solve real world challenges 0:00:23.079 --> 0:00:25.440 and transform the way they do business. 0:00:25.480 --> 0:00:29.560 From accelerating scientific breakthroughs to reimagining education. It's a fresh 0:00:29.600 --> 0:00:33.040 look at innovation in action, where big ideas meet cutting 0:00:33.120 --> 0:00:33.960 edge solutions. 0:00:34.320 --> 0:00:37.160 You'll hear from industry leaders, creative thinkers, and of course 0:00:37.280 --> 0:00:40.920 Malcolm Gladwell himself as he guides you through each story. 0:00:41.440 --> 0:00:44.680 New episodes of Smart Talks with IBM drop every month 0:00:44.680 --> 0:00:48.279 on the iHeartRadio app, Apple Podcasts, or wherever you get 0:00:48.320 --> 0:00:52.240 your podcasts. Learn more at IBM dot com slash smart Talks. 0:00:52.880 --> 0:00:56.200 This is a paid advertisement from IBM. 0:00:56.400 --> 0:00:59.440 In the world of educational research, is a famous video 0:00:59.480 --> 0:01:02.680 of a boy named Sean. I don't mean famous in 0:01:02.680 --> 0:01:04.679 a sense that it has a million views on YouTube. 0:01:05.120 --> 0:01:07.080 I mean that in the circle of people who think 0:01:07.120 --> 0:01:10.640 about teaching and how to make teaching better. The video 0:01:10.680 --> 0:01:14.240 has been written about in journal articles and shown over 0:01:14.280 --> 0:01:17.840 and again in college classrooms. It's a ten minute clip 0:01:18.240 --> 0:01:21.520 of a third grade class somewhere in Michigan. It was 0:01:21.560 --> 0:01:24.280 filmed in January of nineteen ninety, so the video is 0:01:24.319 --> 0:01:27.959 a bit grainy. The teacher's name is Deborah Lowenberg Ball. 0:01:28.319 --> 0:01:31.360 She's a professor at Michigan State University who is part 0:01:31.360 --> 0:01:34.080 of her research. Teaches a one hour math class at 0:01:34.120 --> 0:01:37.680 a local elementary school on the day in question. Miss 0:01:37.760 --> 0:01:41.440 Ball begins by asking her students about the previous day's lesson, 0:01:41.800 --> 0:01:44.600 which was about even and odd numbers. 0:01:44.760 --> 0:01:47.000 I would like to hear from as many people as possible, 0:01:47.400 --> 0:01:49.640 what comments you had, reactions you had to be in 0:01:49.640 --> 0:01:50.560 that meeting yesterday. 0:01:51.480 --> 0:01:54.320 A little boy with black hair raises his hand. His 0:01:54.400 --> 0:01:55.000 name is Sean. 0:01:55.760 --> 0:01:56.000 Hello. 0:01:56.360 --> 0:01:59.120 I don't have anything about the meeting yesterday. 0:01:59.160 --> 0:02:01.880 That was Sean was. 0:02:01.840 --> 0:02:03.480 Thinking about the number six. 0:02:04.120 --> 0:02:06.680 So I was thinking that it's a it's an idd 0:02:06.800 --> 0:02:08.720 it can be an odd number two because there could 0:02:08.760 --> 0:02:17.160 be two, two, four, six, two, three twos and two 0:02:17.400 --> 0:02:26.480 threes and add Antonisina two thinks make Italy two things. 0:02:26.120 --> 0:02:29.800 And Sean doesn't understand what odd and even means. He 0:02:29.919 --> 0:02:32.160 thinks that just because you can break down six in 0:02:32.240 --> 0:02:35.880 an odd number of parts and an even number of parts, 0:02:35.919 --> 0:02:40.000 that six must exist in some magical middle category. And 0:02:40.040 --> 0:02:42.800 when you listen to the Sean videotape, you keep waiting 0:02:42.840 --> 0:02:46.080 for the teacher to say, oh, no, Sean, you misunderstand. 0:02:46.680 --> 0:02:49.640 But Deborah Ball doesn't do that. She never tells him 0:02:49.639 --> 0:02:54.799 he's wrong. Instead, she simply asks him to explain his thinking. 0:02:55.200 --> 0:02:56.960 And the two things that you put together to make 0:02:57.000 --> 0:02:57.760 it were odd right? 0:02:58.440 --> 0:03:02.560 Three and three are each old and thinks Roba so. 0:03:03.639 --> 0:03:07.359 Two or even bald and asked the class to give 0:03:07.400 --> 0:03:10.720 their views. Other students jump up and explain their theories 0:03:10.720 --> 0:03:13.880 on the blackboard. For the next fifteen minutes, she defintely 0:03:13.960 --> 0:03:16.880 guides the class through an in depth investigation of what 0:03:16.960 --> 0:03:21.480 she calls shawn numbers until Sewn himself realizes that the 0:03:21.520 --> 0:03:24.160 real meaning of odd and even is something different than 0:03:24.160 --> 0:03:27.160 he had imagined. And now he gets it. 0:03:27.560 --> 0:03:30.200 I've been great, Thank you for when you are in love. 0:03:31.120 --> 0:03:33.640 I don't want to focus just on how little Sean 0:03:33.960 --> 0:03:36.480 finally made his own way to the right answer. I'm 0:03:36.520 --> 0:03:38.839 interested in what his teacher did to get him there. 0:03:39.600 --> 0:03:43.520 Deborah Ball worked magic. She never told Sean the right answer, 0:03:44.160 --> 0:03:46.200 She just led him to a place where he could 0:03:46.240 --> 0:03:51.120 discover it for himself. My name is Malcolm Glawo. This 0:03:51.240 --> 0:03:54.080 is season six of Smart Talks with IBM, where we 0:03:54.160 --> 0:03:56.720 offer our listeners a glimpse behind the curtain of the 0:03:56.760 --> 0:04:02.000 world of technology and artificial intelligence. In this season, we're 0:04:02.000 --> 0:04:05.200 going to visit companies as varied as Laurel and Ferrari 0:04:05.440 --> 0:04:09.360 and tell stories of how they're using artificial intelligence and 0:04:09.480 --> 0:04:13.800 data to transform the way they do business. This episode 0:04:14.160 --> 0:04:16.520 is about the promise of a radical new idea called 0:04:16.600 --> 0:04:19.919 responsive teaching, the kind of teaching that took place that 0:04:20.080 --> 0:04:24.320 day in Shawn's classroom, and whether artificial intelligence can help 0:04:24.400 --> 0:04:27.200 us train the next generation of teachers to be as 0:04:27.240 --> 0:04:35.640 good as Deborah Ball. Before we talk about how AI 0:04:35.720 --> 0:04:38.400 could transform the way we train teachers, I want to 0:04:38.440 --> 0:04:40.920 go back for a moment to the famous video of Sean. 0:04:41.800 --> 0:04:45.120 In the video, the teacher Deborah Ball doesn't have a 0:04:45.200 --> 0:04:49.640 predetermined plan that she's imposing on the class. She's improvising, 0:04:50.360 --> 0:04:53.760 making up her approach as she goes along, responding to 0:04:53.800 --> 0:04:58.040 her student's odd theory about the number six. Second, she's 0:04:58.080 --> 0:05:03.520 taking Sean seriously. She's not dismissing his theory. She's listening 0:05:03.520 --> 0:05:06.839 to him and trying to understand the problem from his perspective. 0:05:07.600 --> 0:05:11.440 And Thirdly, and most importantly, she's not force feeding him 0:05:11.440 --> 0:05:15.359 the right answer. She's being patient. She's waiting to see 0:05:15.400 --> 0:05:18.440 if with just the right subtle hints, he can get 0:05:18.440 --> 0:05:23.880 to the right answer on his own. Improvisation empathy patients. 0:05:24.720 --> 0:05:26.520 That's responsive teaching. 0:05:26.800 --> 0:05:29.520 What I think about in terms of responsiveness is more like, 0:05:30.640 --> 0:05:33.200 I think that students need to have a sense of 0:05:34.200 --> 0:05:38.560 agency in what happens in the classroom, and like authentic 0:05:38.680 --> 0:05:43.800 agency where they can be legitimized as knowers. 0:05:44.520 --> 0:05:47.920 I spoke to a physicist at Seattle Pacific University named 0:05:47.960 --> 0:05:52.359 Amy Robertson, a longtime advocate for responsive teaching. She uses 0:05:52.400 --> 0:05:54.000 the Sean video in her classroom. 0:05:54.240 --> 0:05:56.440 You have to trust that kids have a way of 0:05:56.520 --> 0:05:59.640 doing that and that, like heard, what she mostly did 0:05:59.720 --> 0:06:02.359 was to facilitate a conversation and to say you have 0:06:02.400 --> 0:06:03.520 to listen to them talk. 0:06:03.920 --> 0:06:06.440 No one told him he was wrong, that's right, and 0:06:06.480 --> 0:06:09.560 then he goes, He goes. I didn't think of it 0:06:09.600 --> 0:06:10.560 that way again. 0:06:10.920 --> 0:06:13.560 I thank you for bringing it up. 0:06:14.000 --> 0:06:17.880 You've expanded my understanding. Thank you for bringing it up again. 0:06:17.920 --> 0:06:20.719 It's like this, I love, I love. 0:06:21.400 --> 0:06:24.680 I Responsive teaching, as I think about it, is kind 0:06:24.720 --> 0:06:28.480 of rooted in this like Eleanor Duckworth's work around the 0:06:28.520 --> 0:06:31.400 Having of Wonderful Ideas, where she says, like, the goal 0:06:31.440 --> 0:06:35.120 of education is for students to have wonderful ideas and 0:06:35.120 --> 0:06:36.520 to have a good time having them. 0:06:36.720 --> 0:06:40.080 But I love that. I've never heard that. What a beautiful, 0:06:40.600 --> 0:06:44.080 succinct way of summing up the purpose of education. Yes, 0:06:45.040 --> 0:06:48.760 responsive teaching is beautiful. It's rare to find a new 0:06:48.800 --> 0:06:52.440 teaching idea that everyone loves. This is one of those 0:06:52.520 --> 0:06:55.680 rare ideas. Watching the Deborah Ball classroom, all I could 0:06:55.680 --> 0:06:58.400 think was, I really really hope my daughters get to 0:06:58.440 --> 0:07:01.760 experience a math class like that. Far too many kids 0:07:01.920 --> 0:07:05.040 are convincing themselves at far too young an age that 0:07:05.120 --> 0:07:08.920 math isn't for them, and responsive teaching is a way 0:07:09.000 --> 0:07:13.200 to solve that problem. But here is the issue. It's 0:07:13.480 --> 0:07:17.360 really really hard to teach responsive teaching. Robertson says that 0:07:17.440 --> 0:07:20.600 teaching exists in a cultural environment where the teacher is 0:07:20.640 --> 0:07:23.720 expected to be the source of truth that teaching is 0:07:23.760 --> 0:07:27.000 about the immediate correction of error and not letting a 0:07:27.080 --> 0:07:31.760 child wander down the pathway of their own misunderstanding. Responsive 0:07:31.800 --> 0:07:36.120 teaching is deeply counterintuitive, and the only way to understand 0:07:36.160 --> 0:07:39.760 its beauty is to do it over and over again. 0:07:40.680 --> 0:07:45.600 Aspiring teachers need a way to practice. For as long 0:07:45.640 --> 0:07:48.480 as there has been technology, people have turned to digital 0:07:48.520 --> 0:07:52.560 machines to solve problems. My father was a mathematician and 0:07:52.600 --> 0:07:55.160 I remember him coming home in the nineteen seventies with 0:07:55.240 --> 0:07:58.280 a big stack of computer cards in his briefcase that 0:07:58.360 --> 0:08:01.840 he used to program the main frame back the office. Today, 0:08:02.120 --> 0:08:06.280 with the rise of artificial intelligence, the scale and complexity 0:08:06.520 --> 0:08:09.720 of the problems technology can help us solve has jumped 0:08:09.760 --> 0:08:12.960 by many orders of magnitude. You must have worked with 0:08:13.000 --> 0:08:16.800 a with a million customers who are experimenting with lll ms. 0:08:16.840 --> 0:08:19.800 Has there been one use case that you were like, WHOA, 0:08:19.880 --> 0:08:23.720 I had no idea or just simply that's clever. I'm 0:08:23.800 --> 0:08:27.480 speaking to Brian Bissel, who works out of IBM's Manhattan office. 0:08:27.880 --> 0:08:31.080 He helps IBM customers discover how best to get AI 0:08:31.200 --> 0:08:32.080 to work for them. 0:08:32.800 --> 0:08:34.480 There is one, but I don't think I can talk 0:08:34.480 --> 0:08:34.800 about it. 0:08:34.880 --> 0:08:38.240 Unfortunately, Wait, wait, you can't tease me like that, can 0:08:38.320 --> 0:08:42.320 you wait? Disguise disguise it for me, Just give me 0:08:42.360 --> 0:08:43.200 a general. 0:08:43.480 --> 0:08:45.280 It was about the ability to pull certain types of 0:08:45.280 --> 0:08:49.000 information out of documents that you you wouldn't think you 0:08:49.000 --> 0:08:52.320 would be able to get the model to do, and 0:08:52.480 --> 0:08:55.440 be able to do that at a very large scale. 0:08:55.800 --> 0:08:58.440 Bissile's point was that we are well past the stage 0:08:58.480 --> 0:09:01.760 where anyone wonders whether a I can be useful. The 0:09:01.800 --> 0:09:04.920 real question now is what problems do we want to 0:09:05.000 --> 0:09:07.720 use it to solve? Where it can make the biggest difference, 0:09:08.440 --> 0:09:11.800 And Basil saw lots of opportunities in education. 0:09:12.720 --> 0:09:15.800 I have two kids, one in middle school and one 0:09:15.880 --> 0:09:19.160 who just graduated high school, and I'm well aware of 0:09:19.240 --> 0:09:22.840 students using things like chat GPT to do their homework. 0:09:23.640 --> 0:09:29.080 And it's very easy to take tools like that and 0:09:29.120 --> 0:09:32.800 even IBM's own large language models and just take a problem, 0:09:33.160 --> 0:09:36.880 a piece of homework, something you want written, and drop 0:09:36.920 --> 0:09:38.800 it into that and have it generate the answer for 0:09:38.880 --> 0:09:43.199 you and the student. The user in that case hasn't 0:09:43.240 --> 0:09:46.240 done any work, they haven't put any real thought into it. 0:09:46.280 --> 0:09:50.240 To Basil, that's the wrong use of AI that's technology 0:09:50.480 --> 0:09:54.280 making is dumber. What we really want is technology that 0:09:54.360 --> 0:09:57.280 makes us smarter. Basil explain to me that there are 0:09:57.320 --> 0:10:00.840 now two big tools being used for AI productivity, AI 0:10:01.000 --> 0:10:06.320 agents and AI assistants. Let's start with AI agents. AI 0:10:06.400 --> 0:10:10.400 agents can reason plan and collaborate with other AI tools 0:10:10.640 --> 0:10:14.640 to autonomously perform tasks for a user. This will gave 0:10:14.679 --> 0:10:17.200 me an example of how college freshmen might use an 0:10:17.200 --> 0:10:17.800 AI agent. 0:10:18.200 --> 0:10:20.720 As a new student, you may not know how do 0:10:20.760 --> 0:10:22.839 I do with my health and wellness issue? So many 0:10:22.880 --> 0:10:25.440 credits are going to get for this given class. You 0:10:25.480 --> 0:10:28.839 could talk to someone and find out some of that, 0:10:29.840 --> 0:10:31.440 but maybe it's a little bit sensitive and you don't 0:10:31.440 --> 0:10:31.959 want to do that. 0:10:32.480 --> 0:10:35.640 Bisill told me you could build an AI agent, a 0:10:35.720 --> 0:10:39.160 resource for new students that helps them navigate a new campus, 0:10:39.480 --> 0:10:42.720 register for classes, access the services they need, and even 0:10:42.720 --> 0:10:46.559 schedule appointments on their behalf, which in turn buys them 0:10:46.559 --> 0:10:48.880 more time to focus on their actual school work. 0:10:49.200 --> 0:10:54.120 We can see patterns of how agents and assistants can 0:10:54.200 --> 0:10:58.599 help employees and customers and end users be more productive. 0:10:58.800 --> 0:11:03.040 Automate workflows are not doing certain types of repetitive work 0:11:03.520 --> 0:11:07.240 over and over again. And streamlining their lives and making 0:11:07.360 --> 0:11:10.880 data more accessible to them twenty four hours a day. 0:11:11.240 --> 0:11:14.960 But Bissil says you can also use AI assistance in 0:11:15.040 --> 0:11:19.160 the education space. AI assistants are reactive as opposed to 0:11:19.200 --> 0:11:23.559 AI agents, which are proactive. AI assistants only performed tasks 0:11:23.600 --> 0:11:28.520 at your request. They're programmed to answer your questions, and 0:11:28.559 --> 0:11:31.640 as it turns out, AI assistants are now being used 0:11:31.960 --> 0:11:35.760 to further the responsive teaching revolution, which is why I 0:11:35.800 --> 0:11:39.280 found myself on a beautiful Georgia spring day not long ago, 0:11:39.440 --> 0:11:43.040 on the campus of Kansas State University, sitting in a 0:11:43.040 --> 0:11:47.200 classroom with two researchers, one of them Professor Dabe Lee. 0:11:47.440 --> 0:11:49.920 Let's go into the journey of building this thing. You 0:11:50.480 --> 0:11:52.040 started w by taking a course. 0:11:52.040 --> 0:11:54.960 What was the course you took, Yeah, so it was 0:11:55.520 --> 0:12:01.959 offered by Coursera. It was designed by IBM AI Foundation 0:12:02.240 --> 0:12:03.040 for everyone. 0:12:06.320 --> 0:12:09.000 In her AI Foundation's course, Lee learned how to build 0:12:09.000 --> 0:12:12.719 an AI assistant using IBM Watson X. That course took 0:12:12.760 --> 0:12:14.000 how long to take. 0:12:14.520 --> 0:12:17.199 It was not to know it was like fourteen weeks. 0:12:18.240 --> 0:12:21.640 Lee's idea was to train an AI assistant on classroom 0:12:21.720 --> 0:12:25.120 data to play the role of sean A digital persona 0:12:25.360 --> 0:12:27.839 of a nine year old who likes math but doesn't 0:12:27.840 --> 0:12:31.840 always understand math, and that AI assistant she thought could 0:12:31.880 --> 0:12:35.000 be used to train preservice teachers or teachers in training 0:12:35.280 --> 0:12:37.880 who are preparing to enter one of the most challenging 0:12:37.880 --> 0:12:39.640 professions in the modern world. 0:12:40.600 --> 0:12:43.600 So when you think about the teacher education and a 0:12:43.800 --> 0:12:49.439 major challenge that teacher education phase is that we need 0:12:49.679 --> 0:12:54.400 children to practice with. We need instructors who will give 0:12:54.520 --> 0:12:59.560 the instruction on the pedagogical skills. So when you look 0:12:59.600 --> 0:13:03.839 at the education program, we have coursework in field experience, 0:13:04.360 --> 0:13:08.720 and in those two areas there is something missing all 0:13:08.800 --> 0:13:09.240 the time. 0:13:10.400 --> 0:13:13.760 Li says that pre service teachers often lack access to 0:13:13.840 --> 0:13:17.360 both students and experienced teachers during their education. 0:13:18.000 --> 0:13:22.400 So what we try to resolve is that we have 0:13:22.520 --> 0:13:25.760 this virtual student for pre service teacher to work with 0:13:26.559 --> 0:13:30.800 so that they can practice their responsive teaching skills. 0:13:31.000 --> 0:13:35.080 The first AI assistant Lee created is g Wu gi Wu, 0:13:35.240 --> 0:13:39.160 emulates the persona of a nine year old third grade girl. Then, 0:13:39.200 --> 0:13:41.880 with the help of one of her collaborators, a researcher 0:13:41.920 --> 0:13:46.199 at Canazon named Sean English, she created two more AI assistants, 0:13:46.600 --> 0:13:51.679 Gabriel and Noah, each of which have their own distinctive characteristics. 0:13:51.920 --> 0:13:55.160 So how are gabriel and Noah different from. 0:13:55.080 --> 0:14:01.120 G Wu Gabrielle? My first one is very short answered. 0:14:01.360 --> 0:14:04.240 If you ask an open ended question, he will answer 0:14:04.280 --> 0:14:08.679 it in a close way. So I use that characteristic. 0:14:08.760 --> 0:14:13.800 And that's the problem that most teachers actually base. They're 0:14:13.840 --> 0:14:17.040 asked children who are shay, who are reserved, and who 0:14:17.040 --> 0:14:21.680 would not share much of their thoughts. So we wanted 0:14:21.760 --> 0:14:25.520 that characteristic in some characters, and we use Gabrielle to 0:14:25.840 --> 0:14:27.000 have that characteristic. 0:14:28.480 --> 0:14:30.960 And Noah. What'snawah's personality? 0:14:32.320 --> 0:14:36.040 How do he playful? Cheery, bright and energetic? 0:14:36.840 --> 0:14:39.600 That's Sean English professor, Lee's fellow researcher. 0:14:40.120 --> 0:14:46.640 And Jewu is articulated and kind of smart, but he 0:14:46.840 --> 0:14:48.840 she has her own way of thinking. 0:14:49.240 --> 0:14:51.160 I would end up spending a lot of time with 0:14:51.280 --> 0:14:54.640 jie Wu. She's something of a character. I asked Sean 0:14:54.680 --> 0:14:57.920 about the process of creating these AI assistants. What does 0:14:58.040 --> 0:15:03.560 building the content side of the AI assistant entail? Sean? 0:15:03.560 --> 0:15:03.600 What? 0:15:04.000 --> 0:15:06.480 It sets up a series of actions, effectively, which are 0:15:07.160 --> 0:15:09.680 response cases. You can kind of think of them as 0:15:09.960 --> 0:15:12.760 you have a series of questions that you tie to 0:15:13.920 --> 0:15:17.720 an intent, and then that intent has reactions from the bot, 0:15:18.040 --> 0:15:21.160 and so effectively, if we were looking to say make 0:15:21.200 --> 0:15:23.520 a hello action, we would have all the different ways 0:15:23.520 --> 0:15:25.880 that people could say Hello, Hello, what's up, how you doing, 0:15:25.880 --> 0:15:26.880 and all that kind of stuff. 0:15:27.240 --> 0:15:31.120 Sean says, the longer the list of potential responses, the better, 0:15:31.600 --> 0:15:35.520 But AI's responses don't just follow the list. The AI 0:15:35.560 --> 0:15:39.120 assistant uses those suggested responses to come up with a 0:15:39.240 --> 0:15:43.120 universe of other responses, and in that process sometimes it 0:15:43.160 --> 0:15:45.480 comes up with things that just don't make sense. 0:15:45.560 --> 0:15:49.040 And from a technological standpoint, while AI is a fantastic tool, 0:15:49.160 --> 0:15:51.600 AI can hallucinate, which I mean, just give things that 0:15:51.600 --> 0:15:54.680 it's just straight up made up. There's a famous example 0:15:54.720 --> 0:15:56.640 of this called the three rs is where you ask 0:15:56.680 --> 0:15:59.640 a popular large language model how many RS are in strawberry, 0:16:00.000 --> 0:16:01.960 and it gives you the wrong answer, and he repeats 0:16:02.000 --> 0:16:04.800 that result repetitively. You always want to have a human 0:16:05.320 --> 0:16:07.320 interacting with the system to be able to go, hey, 0:16:07.920 --> 0:16:10.080 that's a little crazy. I don't think that's exactly what 0:16:10.080 --> 0:16:10.920 we're going for here. 0:16:11.560 --> 0:16:13.560 That's why it's good to have someone like Sean English 0:16:13.560 --> 0:16:15.640 around to step in and get the model back on track, 0:16:16.040 --> 0:16:19.200 and over time, when the model has enough training, it's 0:16:19.280 --> 0:16:25.640 ready for the teachers in training. One of the rollouts 0:16:25.640 --> 0:16:28.720 of Jiwu, Gabriel, and Noah was with the teacher training 0:16:28.760 --> 0:16:30.720 program at the University of Missouri. 0:16:31.040 --> 0:16:33.400 I was just kind of excited to see what the 0:16:33.480 --> 0:16:35.880 program was and what it was going to be doing. 0:16:36.360 --> 0:16:39.400 This is Logan Hovis, a junior at Missouri on the 0:16:39.440 --> 0:16:41.760 path to becoming an elementary school teacher. 0:16:42.160 --> 0:16:44.240 Obviously a little skeptical when he said it was so 0:16:44.520 --> 0:16:47.640 to you know, be like talking to a student. You're like, 0:16:47.640 --> 0:16:50.040 there's no way this AI thing is going to totally 0:16:50.080 --> 0:16:52.760 sound like a second grader or a third grader, Like 0:16:52.800 --> 0:16:55.360 it's going to sound like an adult, or it's going 0:16:55.400 --> 0:16:57.240 to sound like a robot that knows all the answers. 0:16:57.760 --> 0:17:00.480 And it really didn't. It really was like talking to 0:17:00.520 --> 0:17:00.880 a child. 0:17:00.920 --> 0:17:03.480 It was very very well developed in the way that 0:17:03.520 --> 0:17:05.600 you really sit there and you feel like you're talking 0:17:05.600 --> 0:17:06.120 to a kid. 0:17:06.720 --> 0:17:09.920 Her point wasn't that Jiwu and her fellow avatars were 0:17:09.960 --> 0:17:13.720 equivalent to real kids. Of course not, but for someone 0:17:13.840 --> 0:17:16.760 starting out, someone who was already nervous about being plunged 0:17:16.760 --> 0:17:19.480 into a classroom of nine year olds, Jeewu was like 0:17:19.520 --> 0:17:21.400 a warm up before a baseball game. 0:17:21.800 --> 0:17:23.919 What I can think of is like, you know how 0:17:24.320 --> 0:17:26.600 when you're at batting practice for baseball or softball, you 0:17:26.640 --> 0:17:29.560 have those automatic pitchers that throw them because you're working 0:17:29.640 --> 0:17:32.280 on your skill as the hitter. What can I do differently? 0:17:32.280 --> 0:17:35.320 What am I doing wrong? But that doesn't replace the 0:17:35.359 --> 0:17:37.080 game and what you do in a game. But this 0:17:37.200 --> 0:17:40.000 is you getting to practice your own skills to be 0:17:40.000 --> 0:17:41.480 better when you go in a game. And I think 0:17:41.520 --> 0:17:44.600 that's kind of what the AI software feels like for us. 0:17:46.440 --> 0:17:49.119 In batting practice, the pitches don't come as hard and 0:17:49.160 --> 0:17:51.679 fast as the pitch is in a real game, but 0:17:51.720 --> 0:17:53.560 you get to stand at the plate and the pitcher 0:17:53.640 --> 0:17:57.320 throws you dozens of balls over and over again in 0:17:57.359 --> 0:17:59.959 a concentrated block that allows you to work on your 0:18:00.160 --> 0:18:02.760 swing closely and carefully. 0:18:03.320 --> 0:18:05.760 There's a lot less stimulus going on around because the 0:18:05.800 --> 0:18:08.560 classroom is very very busy. It's wonderful, it's beautiful, but 0:18:08.600 --> 0:18:11.080 it's very very busy, so sometimes it's hard to keep 0:18:11.600 --> 0:18:14.880 you know, that focus in on the tasks that they're 0:18:14.920 --> 0:18:17.479 doing at hand, and also in the teacher setting, you're 0:18:17.520 --> 0:18:20.160 also kind of always looking around making sure that other 0:18:20.200 --> 0:18:22.240 students are doing what they're supposed to be doing, but 0:18:22.320 --> 0:18:24.440 also like if they need any help, if everything's going 0:18:24.480 --> 0:18:29.760 okay in the classroom. So being on the Jiwu chat, 0:18:30.240 --> 0:18:32.119 it was just nice that you didn't have to do 0:18:32.280 --> 0:18:34.600 any of the extra work to keep the focus on there, 0:18:35.040 --> 0:18:38.600 and it also felt you did have to feel the 0:18:38.640 --> 0:18:42.240 student's nervousness of being one on one with you, and 0:18:42.440 --> 0:18:44.800 also as a teacher, it was a lot less pressure 0:18:44.840 --> 0:18:47.520 too because I was like, Okay, I'm taking this series. 0:18:47.560 --> 0:18:49.840 This is a student I'm questioning, but. 0:18:49.800 --> 0:18:52.399 I also know I'm probably not going to hurt someone's 0:18:52.400 --> 0:18:55.040 feelings right now, and that's terrifying to think I'm going 0:18:55.080 --> 0:18:58.640 to ask the wrong question and upset the child because 0:18:58.680 --> 0:18:59.479 I've done that. 0:19:00.440 --> 0:19:02.879 We think of the typical use of AI as a 0:19:02.920 --> 0:19:05.680 tool for speeding things up. That's what we always hear 0:19:05.960 --> 0:19:08.879 that the introduction of AI to problem X gave an 0:19:08.880 --> 0:19:13.159 answer in minutes when solving problem X used to take weeks. 0:19:13.720 --> 0:19:17.640 But we shouldn't forget another use that it allows us 0:19:17.960 --> 0:19:21.360 to slow things down. Hoves, if she wanted to, could 0:19:21.359 --> 0:19:24.560 spend a whole weekend practicing with ji Wu. A real 0:19:24.640 --> 0:19:27.320 nine year old will get frustrated on board with the 0:19:27.320 --> 0:19:31.320 fumbling novice after ten minutes, but gi Wu ji Wu 0:19:31.480 --> 0:19:34.320 will happily answer questions for as long as it takes 0:19:34.320 --> 0:19:36.920 for the people who want to learn to be responsive 0:19:37.520 --> 0:19:41.760 to learn how to be responsive. At the end of 0:19:41.800 --> 0:19:45.280 my time at Kennesas State, Sean and Dabe led me 0:19:45.359 --> 0:19:48.320 to a small table where Dabe had set up her laptop. 0:19:48.920 --> 0:19:51.240 In the corner of the screen was a chat box 0:19:51.520 --> 0:19:53.960 of the sort we've all seen and used a thousand times. 0:19:54.720 --> 0:19:58.240 Ji Wu began. She had been given a math problem. 0:19:58.800 --> 0:20:04.400 Rutle, who are of three fourth cup of a flower 0:20:04.720 --> 0:20:09.480 to the ball thanks to the added another three six 0:20:09.680 --> 0:20:14.480 is cup. It's a total amount of flower the use 0:20:15.040 --> 0:20:20.240 greater or dan or a less than one cup? How 0:20:20.320 --> 0:20:21.560 much flower. 0:20:21.280 --> 0:20:22.120 They can use. 0:20:22.520 --> 0:20:25.600 That's a simulation of Giewu speaking. We pause it for 0:20:25.600 --> 0:20:31.080 a second. So Jewu is trying to solve this problem. 0:20:31.160 --> 0:20:33.760 And the first thing she does is she draws a 0:20:33.840 --> 0:20:37.440 rectangle on the screen. This is a common tactic of 0:20:37.480 --> 0:20:41.600 nine year olds try to visualize the fractions. And she 0:20:41.680 --> 0:20:47.040 divides it into four pieces. And now she's gonna color 0:20:47.119 --> 0:20:49.760 in three of the four pieces. Yes, so she's representing 0:20:49.760 --> 0:20:52.760 this is quite good. She's representing three quarters on the screen. 0:20:54.440 --> 0:20:58.080 This is three sixes. 0:21:00.080 --> 0:21:05.520 So now Jiwu does another rectangle with six boxes and 0:21:05.600 --> 0:21:06.840 colors in three of them. 0:21:07.160 --> 0:21:13.960 Okay, together makes sikes going off. 0:21:15.560 --> 0:21:19.520 So then she counts up all the colored boxes and 0:21:19.560 --> 0:21:22.879 that's her numerator, and counts up the total number of 0:21:22.880 --> 0:21:26.520 boxes and that's her denominator. Ji Wu had counted the 0:21:26.560 --> 0:21:30.160 colored boxes and landed on an answer. When you add 0:21:30.400 --> 0:21:33.080 three quarters of a cup and three sixths of a cup, 0:21:33.680 --> 0:21:36.800 you get six tenths of a cup. So, according to 0:21:36.880 --> 0:21:39.800 ji Wu, Martin has less than one cup. And she 0:21:39.920 --> 0:21:41.040 thinks she solved the problem. 0:21:41.200 --> 0:21:43.879 Yes, okay, so it's less than one cup. 0:21:44.200 --> 0:21:47.480 Yeah, so she says it's less than one cup. Now, 0:21:47.560 --> 0:21:50.480 oh my god, this is hard. So the question is 0:21:50.520 --> 0:21:53.880 what do I, as a teacher say to Jiwu. We 0:21:53.880 --> 0:21:57.000 were off. The rules were simple. I couldn't give ji 0:21:57.040 --> 0:21:59.560 Wu the answer or explain to her what she was 0:21:59.560 --> 0:22:02.520 doing wrong. I had to be Deborah Ball. I had 0:22:02.520 --> 0:22:05.879 to help her find the way herself. The chat box 0:22:06.000 --> 0:22:08.040 in the corner of the screen was waiting for my 0:22:08.080 --> 0:22:11.359 first question. I thought for a moment and started typing, 0:22:11.680 --> 0:22:14.720 do you think the boxes in the red rectangle are 0:22:14.760 --> 0:22:17.800 the same size as the boxes in the blue rectangle. 0:22:18.640 --> 0:22:21.040 Then I turned to Sean and dabey, is that a 0:22:21.040 --> 0:22:21.560 good question? 0:22:21.800 --> 0:22:23.280 Yeah, serious thing. 0:22:24.280 --> 0:22:25.719 Yeah, that's a good question. 0:22:26.240 --> 0:22:30.359