WEBVTT - Khan Academy's AI Pilot Program 0:00:00.080 --> 0:00:03.920 These sees Bloomberg Business Week with Carol Messer and Tim 0:00:03.960 --> 0:00:07.640 Stenebec on Bloomberg Radio. We did catch up with sal 0:00:07.720 --> 0:00:10.559 Cohn in late January. We got his thoughts on learning 0:00:10.560 --> 0:00:12.920 in our post pandemic world, what things are sticking, what 0:00:13.039 --> 0:00:16.200 things are not. We talked online education, something he and 0:00:16.200 --> 0:00:18.279 his team have been doing for about fifteen years or so, 0:00:19.440 --> 0:00:21.639 and now educators of all types, as you know, have 0:00:21.680 --> 0:00:24.560 a new tool maybe to use. So back to talk 0:00:24.600 --> 0:00:26.360 with us once again about the role of AI and 0:00:26.520 --> 0:00:30.240 education and their specific moves in that regard. So com 0:00:30.240 --> 0:00:33.080 back with us. Founder and CEO Khan Academy with us 0:00:33.159 --> 0:00:35.479 via zoom from Mountain View, California. Sow, good to have 0:00:35.520 --> 0:00:38.280 you back with us. How are you. I'm doing great, 0:00:38.280 --> 0:00:39.839 Thanks for having me. Well, it's great to have you 0:00:39.920 --> 0:00:42.879 here once again. Late January. We talked about a lot 0:00:42.920 --> 0:00:45.560 of things, including AI, which it feels like all of 0:00:45.600 --> 0:00:47.760 a sudden everybody was talking about it once again. You're 0:00:47.840 --> 0:00:51.920 launching your own AI platform. Can you tell us about it? Yeah, 0:00:52.000 --> 0:00:54.720 And when we talked in January, I was under n 0:00:55.280 --> 0:00:59.040 I was under NDA, so I couldn't tell you what. Okay. 0:00:59.440 --> 0:01:03.480 Back in back in August, open AI reached out to us. 0:01:04.040 --> 0:01:06.440 They were working on GPT four. As many folks know 0:01:06.560 --> 0:01:08.640 Chat GPT when it came out in November, it was 0:01:08.640 --> 0:01:12.040 based on GPT three or three point five. But when 0:01:12.040 --> 0:01:15.320 we saw what was possible with GPT four, we said, hey, 0:01:15.360 --> 0:01:17.640 this could actually allow us to address some of the 0:01:17.680 --> 0:01:21.520 holy grails in education, making it truly interactive, giving every 0:01:21.520 --> 0:01:24.880 student a socratic tutor, being able to do things that 0:01:25.040 --> 0:01:28.240 might have only looked like science fiction a few months ago. 0:01:28.600 --> 0:01:32.480 And Open AI wanted to work with us because they said, look, 0:01:32.880 --> 0:01:35.280 GPT four is going to be exciting but also scary 0:01:35.680 --> 0:01:37.560 in certain ways. We want to be able to launch 0:01:37.600 --> 0:01:41.120 with some social positive use cases, especially in education, especially 0:01:41.200 --> 0:01:44.760 from folks that with folks that people trust. So we've 0:01:44.800 --> 0:01:48.440 been working pretty feverishly on it for many months under NDA. 0:01:48.480 --> 0:01:50.280 So it was really hard in January when you were 0:01:50.320 --> 0:01:52.720 asking me questions about Chat GPT for me to not 0:01:52.800 --> 0:01:56.240 spill the beads. But a couple of weeks ago we 0:01:56.360 --> 0:02:00.440 launched and what we're calling con Migo, which is really 0:02:01.440 --> 0:02:06.320 our incarnation of AI on Kin Academy using GPT four. 0:02:06.760 --> 0:02:10.760 It does essentially two major strands of things. One is 0:02:10.800 --> 0:02:13.880 now when a student is doing something on kin Academy, 0:02:14.200 --> 0:02:17.000 it acts as their tutor, and it really does as 0:02:17.000 --> 0:02:19.200 opposed to just giving you the answer. If a student says, 0:02:19.639 --> 0:02:21.519 you know, telling me the interests, I'm your tutor. I 0:02:21.520 --> 0:02:23.320 can help you and if you know, what do you 0:02:23.360 --> 0:02:24.960 think is the next step. So it really tries to 0:02:24.960 --> 0:02:27.200 push them socratically. It's not just in math, it's in 0:02:27.280 --> 0:02:32.320 every subject. We have history, science, reading, comprehension. There's now 0:02:32.400 --> 0:02:35.080 a tutor and this is a beta now where we've 0:02:35.080 --> 0:02:37.400 only launched it to a few thousand people. There's a 0:02:37.440 --> 0:02:40.080 waiting lists that's forming for folks who want to join. 0:02:40.120 --> 0:02:42.000 It's about twenty thousand people right now, so we're trying 0:02:42.040 --> 0:02:44.840 to take people off as quickly as possible. But the 0:02:44.880 --> 0:02:47.200 other thing we realized is the things that would have 0:02:47.200 --> 0:02:51.040 been impossible without AI. We have activities now where students 0:02:51.040 --> 0:02:56.000 can have conversations with historical characters or literary characters. There 0:02:56.080 --> 0:02:58.919 was a student recently, a high school student. She was 0:02:58.960 --> 0:03:01.119 trying to understand some of the symbolism and the Great 0:03:01.160 --> 0:03:03.080 Gatsby and she says, wait, I have conmigo, I can 0:03:03.240 --> 0:03:05.560 I can now talk to Jay Gatsby himself? And she 0:03:05.639 --> 0:03:08.400 has why are you staring at the green light? And 0:03:08.400 --> 0:03:11.040 he's like, well, it symbolizes the things that I'm trying 0:03:11.080 --> 0:03:13.840 to attain but I just can't. And she actually apologized 0:03:13.840 --> 0:03:16.680 to him for taking up this time. The AI simulation. 0:03:17.320 --> 0:03:19.840 We're creating teacher tools for them to create lesson plans 0:03:19.880 --> 0:03:22.360 for them to refresh their own knowledge. We have a 0:03:22.520 --> 0:03:24.920 kind of a built in guidance counselor now as an 0:03:24.960 --> 0:03:28.359 AI I see a quick quick take piece. Yeah, a 0:03:28.440 --> 0:03:31.079 conversation between teachers and students. We're going to be talking 0:03:31.200 --> 0:03:34.320 and con migo. Yeah. Yeah, But it's just so it's 0:03:34.440 --> 0:03:37.280 it's fascinating because of the amount of use cases that 0:03:37.320 --> 0:03:41.840 you mentioned. Is it available? Talk to me about the availability, 0:03:41.840 --> 0:03:46.080 because I don't think I'm clear on how widespread access 0:03:46.200 --> 0:03:48.000 is at this moment and where you're at in that 0:03:48.120 --> 0:03:52.600 and where it's going. Yeah. So when we launched two 0:03:52.600 --> 0:03:54.880 weeks ago, and we're still kind of in the process 0:03:54.920 --> 0:03:57.240 of launch launching, there's a couple of things we wanted 0:03:57.240 --> 0:03:58.800 to figure out. If one, we just wanted to make 0:03:58.840 --> 0:04:03.640 sure it's working well for in the last weeks we've 0:04:03.680 --> 0:04:05.960 started to get some really positive feedback on that one. 0:04:06.120 --> 0:04:07.440 The other thing we're trying to figure out is how 0:04:07.440 --> 0:04:09.240 do we cost this out and how do we resource 0:04:09.360 --> 0:04:13.280 this the computation costs for leveraging these these next generation 0:04:13.360 --> 0:04:15.880 models like GPT four, it is not free even if 0:04:15.880 --> 0:04:19.480 you're a Chat GPT pro customer. And these are folks 0:04:19.480 --> 0:04:21.840 paying twenty dollars a month to open AI. They only 0:04:21.880 --> 0:04:24.839 get GPT four for I think at twenty five interactions 0:04:25.440 --> 0:04:27.480 per day and you have to wait three hours, so 0:04:27.520 --> 0:04:29.920 they've so in a lot of ways, Conmigo, what we're 0:04:29.920 --> 0:04:33.839 doing is in many ways the most unfettered access to 0:04:33.880 --> 0:04:37.120 GPT four. So we're trying to figure out ways that 0:04:37.880 --> 0:04:39.800 we can we can resource it. So right now there's 0:04:39.800 --> 0:04:42.200 a waiting list, when people to get off the waiting list. 0:04:42.279 --> 0:04:44.640 Right now we're prioritizing people who are donating. It's a 0:04:44.680 --> 0:04:47.920 twenty dollars donation. It helps us resource the computation costs 0:04:47.960 --> 0:04:49.479 and also do the R and D. We are not 0:04:49.560 --> 0:04:52.719 for profits, so that is a donation, a true donation. 0:04:53.160 --> 0:04:54.640 But we're trying to figure out how we can bring 0:04:54.720 --> 0:04:57.680 that that that price down over time so that we 0:04:57.720 --> 0:05:00.919 can make it more and more accessible. Because rather I 0:05:01.000 --> 0:05:03.599 think about you and kon Academy as this great leveler 0:05:03.720 --> 0:05:06.560 when it comes to education. Right, your videos are out there, 0:05:06.560 --> 0:05:08.880 anybody can access them. So how do we make sure 0:05:08.920 --> 0:05:13.080 that once again, we don't have a technology potentially that 0:05:13.440 --> 0:05:17.200 widens the gap rather than reduces it. Yeah, this is 0:05:17.240 --> 0:05:19.480 front of mine for us, and I think there's two dimensions. 0:05:19.480 --> 0:05:21.440 So that one is just the cost issue that we've 0:05:21.440 --> 0:05:24.840 already talked about. Obviously, everything that we stand for is 0:05:25.160 --> 0:05:27.919 our mission is free world class education for anyone anywhere. 0:05:28.160 --> 0:05:31.360 We're going to try to make this as accessible as possible. 0:05:31.680 --> 0:05:35.520 Keeping in mind there is a higher cost, but if 0:05:35.520 --> 0:05:37.800 you think about if we can bring this down to 0:05:38.480 --> 0:05:40.760 let's say a few dollars per student per month, this 0:05:40.839 --> 0:05:45.720 is dramatically lower cost than conventional tutoring. And a lot 0:05:45.760 --> 0:05:47.359 of the school districts have been spending a lot of 0:05:47.360 --> 0:05:50.360 money on conventional tutoring lately with money that was released 0:05:50.360 --> 0:05:54.480 after the pandemic, and they haven't seen results because it 0:05:54.520 --> 0:05:56.680 hasn't been connected to what students are doing in class, 0:05:56.720 --> 0:05:59.880 and it isn't happening during class, So the kids who 0:05:59.880 --> 0:06:03.039 are engaging or are usually the kids who don't need it. 0:06:03.400 --> 0:06:05.520 The kids who need it aren't engaging after school, etc. 0:06:06.080 --> 0:06:07.560 So we think we're gonna be able to get this 0:06:07.680 --> 0:06:10.880 to a much broader set of students. We're starting to 0:06:10.920 --> 0:06:14.000 partner with some school districts, large urban school districts, so 0:06:14.040 --> 0:06:17.640 we can bring it to their students for free, and 0:06:17.680 --> 0:06:21.280 then hopefully we make it more and more accessible over time. 0:06:21.320 --> 0:06:23.520 The other dimension to your question, the way we've been 0:06:23.560 --> 0:06:26.080 thinking about it is a lot of school districts announce 0:06:26.120 --> 0:06:30.080 that they're banning chat GPT, and there's some good reason 0:06:30.160 --> 0:06:32.839 why they're doing that. Obviously, students can use it to cheat, 0:06:32.880 --> 0:06:35.120 They can do shady things with it because it's a 0:06:35.200 --> 0:06:40.080 very open tool on Conmigo, all of the conversations are 0:06:40.160 --> 0:06:43.760 viewable by the teachers and the parents. There's a strong moderation. 0:06:44.600 --> 0:06:47.960 All of the activities are designed to be very pedagogically positive. 0:06:48.040 --> 0:06:51.239 There's hopefully no cheating. More. It will write a paper 0:06:51.279 --> 0:06:52.880 with you, it won't write a paper for you. It 0:06:52.920 --> 0:06:55.440 will tutor you, but it will not tell you the answer. 0:06:56.000 --> 0:06:58.440 So some of those same school districts who banned it, 0:06:58.720 --> 0:07:00.680 and I was afraid when the band because I'm like, 0:07:00.680 --> 0:07:02.680 this is going to create another digital divide because the 0:07:03.160 --> 0:07:05.560 well resource kids are still going to use it. Those 0:07:05.600 --> 0:07:07.920 same school districts who banned it. Now we're saying, hey, 0:07:07.960 --> 0:07:09.520 this is what we wanted. We don't want to stick 0:07:09.520 --> 0:07:11.440 our head in the sand. We want to move forward 0:07:11.480 --> 0:07:12.800 with the times. We just want to do it in 0:07:12.800 --> 0:07:16.360 a safe way that's actually pedagogically positive. So we're pretty 0:07:16.360 --> 0:07:19.560 excited that hopefully we can bring get the cost down. 0:07:19.600 --> 0:07:22.600 Some of that is open AI and Azure and Microsoft 0:07:22.600 --> 0:07:25.760 and all of them getting the efficiencies, and then on 0:07:25.800 --> 0:07:28.120 our side, we're trying to bring the efficiencies in so 0:07:28.160 --> 0:07:30.320 that we can make it as accessible as possible and 0:07:30.560 --> 0:07:33.080 make sure that all kids have access to this. But 0:07:33.200 --> 0:07:36.840 access is such a multi layered thing, Sal, I wonder 0:07:37.040 --> 0:07:39.520 how you're thinking about partnerships when it comes to that, 0:07:39.560 --> 0:07:42.400 because we're still trying to wire up the country. Well, right, 0:07:42.440 --> 0:07:45.280 we've got the internet access thing. I'm on your website. 0:07:45.280 --> 0:07:48.720 It looks like it's like an iPhone is being used 0:07:48.880 --> 0:07:51.760 or samsuent or whatever it is. Like, what about people 0:07:51.760 --> 0:07:54.360 who don't have access to a smartphone? Right, we have 0:07:54.400 --> 0:07:56.160 about a minute and a half left, But I wonder 0:07:56.200 --> 0:07:57.800 how you're thinking about that. And then we'll come back 0:07:57.800 --> 0:08:02.200 and talk some more. Yeah, it's a tough question. Simple 0:08:02.240 --> 0:08:05.120 answer is digital access in the US at least has 0:08:05.120 --> 0:08:08.080 gotten dramatically better. Cell Phone access in the rest of 0:08:08.120 --> 0:08:10.760 the world also has gotten better. We are trying to 0:08:10.800 --> 0:08:13.800 skate to where the puck is going on that front. 0:08:14.720 --> 0:08:17.840 In the US, we think it's all about getting it 0:08:17.880 --> 0:08:20.760 to students in the school setting because at home, as 0:08:20.760 --> 0:08:23.960 you know, there is more inconsistent access. And so that's 0:08:23.960 --> 0:08:27.080 why we're working so closely with school districts, which for 0:08:27.120 --> 0:08:29.840 the most part now do have one to one laptops 0:08:29.920 --> 0:08:32.960 or at least close to that inside of their classrooms. 0:08:33.120 --> 0:08:35.040 And then in other countries, in places like India, we're 0:08:35.040 --> 0:08:38.440 partnering with state governments and that is multi multi layered. 0:08:38.440 --> 0:08:40.720 We need to translate the content, we need to align 0:08:40.720 --> 0:08:42.200 it to their standards, and we also have to figure 0:08:42.240 --> 0:08:44.400 out how we get the vice access right. This is 0:08:44.400 --> 0:08:45.920 why we always like talking to you because it's the 0:08:45.920 --> 0:08:48.680 global perspective. It's not just about teachers and kids and 0:08:48.679 --> 0:08:50.840 here in the US, but it's really about teachers and 0:08:50.920 --> 0:08:53.720 kids everywhere in the world. So so one of the 0:08:53.720 --> 0:08:56.440 things you're talking about, you know, your new AI platform 0:08:56.440 --> 0:08:59.640 that you guys have created, and I came across, are 0:08:59.679 --> 0:09:03.439 we and prepping for our conversation with you. Stanford put 0:09:03.440 --> 0:09:05.600 out a report earlier this month warning that decisions made 0:09:05.600 --> 0:09:09.360 by AI could lead to nuclear level catastrophe. Why is 0:09:09.360 --> 0:09:12.400 it not too early to start playing around with the 0:09:12.480 --> 0:09:18.160 next level AI when it comes to education? Well, I think, 0:09:18.200 --> 0:09:21.120 like any technology, there could be a lot of positives 0:09:21.120 --> 0:09:23.880 and there could also be a lot of possible negatives. 0:09:24.000 --> 0:09:26.680 And I think as a society, it's all about mitigating 0:09:26.679 --> 0:09:29.920 the negatives and then using it for the positives where possible. 0:09:31.760 --> 0:09:34.400 The genie is to some degree out of the bottle 0:09:34.480 --> 0:09:36.400 a little bit, and so I know, I tell this 0:09:36.520 --> 0:09:38.559 the team at Kon Academy, and tell this to our donors. 0:09:38.760 --> 0:09:41.880 I view it as our responsibility as a nonprofit in 0:09:42.080 --> 0:09:46.599 education to show how you can use these technologies and 0:09:46.679 --> 0:09:49.600 a very positive and a very safe and a very 0:09:49.640 --> 0:09:53.800 trusting way, Because if we do that, then we're going 0:09:53.840 --> 0:09:55.360 to be able to reach all these students that we've 0:09:55.400 --> 0:09:57.920 been talking about, and we'll hopefully be able to accelerate 0:09:57.920 --> 0:10:02.960 their learning pretty dramatically. Now, there are other concerns about 0:10:03.160 --> 0:10:05.559 what happens if you have super intelligent AIS. You know, 0:10:05.559 --> 0:10:08.040 who know who knows what GPT five or six or 0:10:08.080 --> 0:10:11.000 seven is going to be able to do planning military 0:10:11.040 --> 0:10:14.360 strategy or figuring out what targets are, etc. But even there, 0:10:15.280 --> 0:10:19.440 I personally think that I would rather have the actors 0:10:19.440 --> 0:10:22.199 in the US being on the cutting edge than US 0:10:22.280 --> 0:10:27.040 pausing and slowing things down and that and letting folks 0:10:27.080 --> 0:10:30.360 who are maybe not as altruistic continue to move forward. 0:10:31.559 --> 0:10:33.320 You know, no matter what we do, I guarantee you 0:10:33.360 --> 0:10:35.080 the Chinese or the Russians are not going to be 0:10:35.120 --> 0:10:38.040 stopping their generative large language model research. So I actually 0:10:38.080 --> 0:10:41.720 think it's a national security issue to keep moving, not 0:10:41.800 --> 0:10:44.960 to stop. Yeah, that's that's such an incredible point. And 0:10:45.000 --> 0:10:47.320 also that you were mentioning that this is just a 0:10:47.360 --> 0:10:50.400 story we've heard before, Like every single innovation has this 0:10:50.480 --> 0:10:55.840 big question, and you can't just like let the United 0:10:55.840 --> 0:10:58.440 States be the only one to not have social media, 0:10:58.480 --> 0:11:01.920 for example, because we're worried about the potential downfalls. Right, 0:11:03.720 --> 0:11:07.439 that's right, And you know social media is not I don't. Well, 0:11:08.040 --> 0:11:10.680 it can be a national security issues, as we've sometimes seen. 0:11:10.720 --> 0:11:13.080 It can lead to polarization, it can lead to you know, 0:11:13.120 --> 0:11:17.240 people believing things that aren't true, but it doesn't quite 0:11:17.320 --> 0:11:20.880 have the same strategic consequence as something like artificial intelligence. 0:11:20.960 --> 0:11:24.120 So an artificial intelligtion is going to intersect with social media, 0:11:24.120 --> 0:11:26.120 It's going to be influencing how people think, et cetera. 0:11:26.240 --> 0:11:30.360 So I feel much better if relatively positive actors in 0:11:31.040 --> 0:11:35.600 the States and aligned or countries aren't thinking about how 0:11:35.640 --> 0:11:39.400 to use this responsibly and not falling behind folks who 0:11:39.400 --> 0:11:43.360 are who have less less issues with using it irresponsibly. 0:11:43.600 --> 0:11:47.839 Are you thinking about or concerned about any legislation on AI. 0:11:47.880 --> 0:11:51.000 I know we talked about rulings at the school district level, 0:11:51.040 --> 0:11:53.760 but are you concerned about any federal legislation that might 0:11:53.800 --> 0:11:59.959 impact usage. I'm always worried when people want to pass legislation, 0:12:00.000 --> 0:12:02.520 should a regulation without having a clear picture of what 0:12:02.559 --> 0:12:05.320 they're trying to protect against. I think right now there's 0:12:05.320 --> 0:12:08.480 a lot of knee jerk fears based on dystopian science 0:12:08.520 --> 0:12:11.960 fiction novels about AIS taking over, etc. You know, a 0:12:11.960 --> 0:12:15.280 lot of the folks calling for regulation can't exactly pinpoint 0:12:15.320 --> 0:12:18.800 to what tangibly they think is going to happen, and 0:12:19.240 --> 0:12:23.400 they can't point to actual examples of that happening. You know, 0:12:23.440 --> 0:12:26.040 the Internet is actually much more of a wild West 0:12:26.320 --> 0:12:29.880 than large language models are right now, and I think 0:12:29.920 --> 0:12:32.679 for the most part, it's probably a good thing. It's 0:12:32.720 --> 0:12:36.440 been left unregulated. I'm not antiregulation. What I would say is, 0:12:36.679 --> 0:12:41.200 let's see where there are issues that pop up, and 0:12:41.240 --> 0:12:44.160 then determine is regulation the best mechanism to do that, 0:12:44.480 --> 0:12:46.040 and then and then that should be a you know, 0:12:46.200 --> 0:12:50.040 chev robust conversation about that. For teachers, what would something 0:12:50.080 --> 0:12:54.040 like this, this kind of functionality and using generative AI 0:12:54.200 --> 0:12:56.880 specifically like you guys are doing, What does it mean 0:12:56.920 --> 0:12:59.800 for teachers specifically who We recently had an author on 0:13:00.160 --> 0:13:03.000 an incredible book and just you know, she embedded herself 0:13:03.040 --> 0:13:06.080 for a year with three different teachers and the stresses, 0:13:06.120 --> 0:13:08.960 the strains not paid well. It's it's a really tough 0:13:09.040 --> 0:13:11.880 profession right now, and to be fair, not really well 0:13:11.920 --> 0:13:16.080 respected by many people out there. And that's unfortunate because 0:13:16.080 --> 0:13:19.720 this is talking about educating the next generation. So what 0:13:20.000 --> 0:13:25.480 might something like this do for teachers? Well, first and foremost, 0:13:25.840 --> 0:13:28.160 you know, your average teacher has thirty kids in the classroom. 0:13:28.200 --> 0:13:31.120 They wish they could replicate themselves, give more personalized attention 0:13:31.120 --> 0:13:33.680 to every student. This is something we've always worked on 0:13:33.720 --> 0:13:36.720 kind academy, but having this AI tutor in the room 0:13:36.760 --> 0:13:40.200 with you. It is like replicating you, we've already seen this. 0:13:40.240 --> 0:13:44.079 When we've been piloting it at some classrooms, the teacher's like, wow, 0:13:44.120 --> 0:13:45.800 it just answered a question that I would have had 0:13:45.840 --> 0:13:49.440 to try to answer, but I wouldn't been able to 0:13:49.440 --> 0:13:51.800 connect with this other student. So it's already doing that. 0:13:51.880 --> 0:13:55.160 But I think even more importantly, teachers spend about thirty 0:13:55.160 --> 0:14:01.520 forty fifty percent of their time planning less than refreshing knowledge. 0:14:02.200 --> 0:14:05.400 We have activities four teachers. We've made them a first 0:14:05.480 --> 0:14:08.959 year user here where it helps them develop lesson plans 0:14:09.000 --> 0:14:12.120 in seconds that would have otherwise have taken hours. It 0:14:12.160 --> 0:14:15.520 helps them prepare their knowledge, It helps them think about 0:14:15.600 --> 0:14:18.080 like almost like an instructional coach, what they how they 0:14:18.080 --> 0:14:21.000 can intervene with their students. So we're kind of calling 0:14:21.080 --> 0:14:24.000 Kanmigo the r AI a tutor for every student and 0:14:24.040 --> 0:14:26.320 a teaching assistant or an army of teaching assistance for 0:14:26.360 --> 0:14:29.920 every teacher. And so I think it'll hopefully lower their 0:14:30.360 --> 0:14:32.120 burden and be able to free up more time for 0:14:32.200 --> 0:14:34.840 the teachers to spend with the students. Can it help 0:14:34.920 --> 0:14:39.720 with grading at all the simpleasures? Yes, I think it is. 0:14:40.280 --> 0:14:43.240 We didn't launched two weeks ago. We have been piloting 0:14:43.240 --> 0:14:44.720 some of this, but we want to make sure that 0:14:44.760 --> 0:14:48.240 it's really robust before we launch it. But I think 0:14:48.280 --> 0:14:51.160 within the next year you're going to see things like 0:14:51.640 --> 0:14:56.480 a teacher saying, hey, Kanmigo, let's develop an activity for 0:14:56.520 --> 0:15:00.440 our students, and Kanmigo says, all right, let's have them 0:15:00.480 --> 0:15:02.680 all right, you know, whether the Civil War could have 0:15:02.680 --> 0:15:04.600 been avoided and how, and then the teachers like that 0:15:04.640 --> 0:15:08.000 sounds great, And then every student sees that assignment and 0:15:08.040 --> 0:15:10.800 then they work with Conmigo to kind of talk about 0:15:12.240 --> 0:15:14.680 whether the Civil War would have happened otherwise. And then 0:15:14.880 --> 0:15:17.840 Kanmigo assesses it and tells the teacher, Hey, these students 0:15:17.840 --> 0:15:22.040 had some really exemplary responses. These students struggled. You know, 0:15:22.360 --> 0:15:24.520 I would give these students an a. I would have 0:15:24.560 --> 0:15:26.560 these students continue to work on it. Right, that's not 0:15:26.640 --> 0:15:28.760 science fiction anymore. I think we're months away from that, 0:15:28.840 --> 0:15:31.560 not years, Especially when we talk about individualized learning, like 0:15:31.760 --> 0:15:34.280 you think about this, it sounds conceptually like it could 0:15:34.280 --> 0:15:36.280 really take us to that level. Hey, Saal just got 0:15:36.280 --> 0:15:39.040 about twenty five thirty seconds. Well we look back five 0:15:39.120 --> 0:15:41.440 years from now, ten years, and is there a superlative 0:15:41.480 --> 0:15:45.240 to describe what generative AI may mean to education. And 0:15:45.280 --> 0:15:47.520 I'm not trying to look for that crazy headline, but 0:15:47.560 --> 0:15:50.320 I'm just curious how you're looking at it. Just quickly. 0:15:51.560 --> 0:15:54.440 I think this is the biggest inflection point of our lifetimes, 0:15:54.520 --> 0:15:56.200 and I think this is going to give a tutor 0:15:56.240 --> 0:15:58.560 to every student on the planet and a teaching assistant 0:15:58.560 --> 0:16:01.440 to every teacher. So it's a big deal. Well, listen, 0:16:01.480 --> 0:16:03.680 thank you so much, and so appreciate you coming back 0:16:03.720 --> 0:16:06.160 when you couldn't talk about it, now you could. We 0:16:06.200 --> 0:16:09.520 really appreciate you sharing that with us in our Bloomberg 0:16:09.520 --> 0:16:13.200 listeners and viewers. So appreciate it. All right. SoCon, be well, 0:16:13.240 --> 0:16:16.320 Take care. SoCon, Founder, chief executive officer of Khan Academy.