WEBVTT - The Future of Learning in the AI Era

0:00:02.480 --> 0:00:06.840
<v Speaker 1>Bloomberg Audio Studios, Podcasts, radio news.

0:00:07.160 --> 0:00:11.040
<v Speaker 2>This is Bloomberg Business Week with Carol Messer and Tim

0:00:11.080 --> 0:00:13.640
<v Speaker 2>Steneveek on Bloomberg Radio.

0:00:15.040 --> 0:00:15.360
<v Speaker 3>Tech.

0:00:16.960 --> 0:00:21.400
<v Speaker 4>Good children, Well if others have.

0:00:24.000 --> 0:00:26.799
<v Speaker 2>Well, our next guest over the last fifteen years or

0:00:26.800 --> 0:00:29.120
<v Speaker 2>so has helped countless students who needed a bit of

0:00:29.160 --> 0:00:33.040
<v Speaker 2>assistance outside of the classroom with math, economics, physics, history

0:00:33.080 --> 0:00:35.280
<v Speaker 2>and more. Maybe it was for a bit of help

0:00:35.320 --> 0:00:37.800
<v Speaker 2>with ap calculus in high school, as it was for

0:00:37.840 --> 0:00:40.919
<v Speaker 2>our producer Elizabeth Cedrin, or for me help with understanding

0:00:40.960 --> 0:00:43.040
<v Speaker 2>accounting principles in business school.

0:00:43.200 --> 0:00:44.839
<v Speaker 3>I love it. Everybody has tapped into it.

0:00:44.880 --> 0:00:46.960
<v Speaker 1>We just had a colleague talk to us about what

0:00:47.280 --> 0:00:50.120
<v Speaker 1>he is doing and tapping into it. And it's not

0:00:50.240 --> 0:00:52.440
<v Speaker 1>someone who's like ten or twelve or fifteen.

0:00:52.800 --> 0:00:55.000
<v Speaker 2>No, he's like downloading the algebra one.

0:00:54.960 --> 0:00:57.160
<v Speaker 1>Right to resprect me cool thinking about it all right.

0:00:57.280 --> 0:01:00.560
<v Speaker 1>Selcon a familiar name and voice to so many people

0:01:00.560 --> 0:01:00.840
<v Speaker 1>out there.

0:01:00.800 --> 0:01:02.160
<v Speaker 3>Who've seen his YouTube videos.

0:01:02.360 --> 0:01:04.479
<v Speaker 1>He is the founder and see of the nonprofit con

0:01:04.840 --> 0:01:08.040
<v Speaker 1>Academy as well as the con Lab School. He's got

0:01:08.080 --> 0:01:12.160
<v Speaker 1>a new book out, Brave New Words. How AI will

0:01:12.200 --> 0:01:16.040
<v Speaker 1>revolutionize education? Why that's a good thing? So it is

0:01:16.600 --> 0:01:18.880
<v Speaker 1>so good to have you here with us, and you know,

0:01:18.920 --> 0:01:20.839
<v Speaker 1>it's funny you just we've been talking a lot about

0:01:20.880 --> 0:01:23.959
<v Speaker 1>artificial intelligence as you as you can imagine. First of all,

0:01:23.959 --> 0:01:25.640
<v Speaker 1>how are you tell us about your world and how

0:01:25.640 --> 0:01:26.160
<v Speaker 1>things have been?

0:01:27.880 --> 0:01:30.399
<v Speaker 4>Well, thanks for having me. I'm great. Things are busy,

0:01:30.680 --> 0:01:33.720
<v Speaker 4>but busy in a good way. It's exciting, you know.

0:01:33.720 --> 0:01:36.679
<v Speaker 4>Obviously the book just got released, But more of what

0:01:36.720 --> 0:01:39.280
<v Speaker 4>the book's about is we're living in a science fiction

0:01:39.360 --> 0:01:42.240
<v Speaker 4>book these days. Con Academy, a lot of folks know.

0:01:43.560 --> 0:01:45.240
<v Speaker 4>I s always great to hear that even people in

0:01:45.240 --> 0:01:47.600
<v Speaker 4>your own team are benefiting from it or have benefited

0:01:47.640 --> 0:01:49.919
<v Speaker 4>from it. But it all started with me tutoring my cousin,

0:01:50.040 --> 0:01:52.120
<v Speaker 4>and everything that we've been doing over the last twenty

0:01:52.200 --> 0:01:53.960
<v Speaker 4>years since then has been trying to try to scale

0:01:54.000 --> 0:01:56.720
<v Speaker 4>up that type of personalization and what we're now seeing

0:01:56.720 --> 0:01:59.040
<v Speaker 4>with generative AI. Even though there's a lot of issues,

0:01:59.080 --> 0:02:00.800
<v Speaker 4>a lot of risks, a lot of things to work through,

0:02:00.840 --> 0:02:03.360
<v Speaker 4>and I try to address that in the book, there's

0:02:03.400 --> 0:02:05.200
<v Speaker 4>a lot of optimism, or at least I have a

0:02:05.240 --> 0:02:07.320
<v Speaker 4>lot of hope that it can can do some really

0:02:07.360 --> 0:02:10.040
<v Speaker 4>positive things, getting us that much closer to giving world

0:02:10.080 --> 0:02:11.280
<v Speaker 4>class education to everyone.

0:02:11.520 --> 0:02:14.239
<v Speaker 2>Well, it's so interesting, Saul, because when we talk about

0:02:14.320 --> 0:02:16.880
<v Speaker 2>AI in the classroom, I think a year ago we

0:02:16.880 --> 0:02:19.480
<v Speaker 2>were talking more about this, this idea that people were

0:02:19.520 --> 0:02:24.000
<v Speaker 2>so professors, teachers were so concerned about chet GPT, for example,

0:02:24.080 --> 0:02:27.800
<v Speaker 2>being used to write term papers or write essays, and

0:02:28.160 --> 0:02:30.400
<v Speaker 2>you know, having the software to detect whether or not

0:02:30.440 --> 0:02:35.400
<v Speaker 2>a GPT wrote an essay the question about having phones

0:02:35.400 --> 0:02:38.200
<v Speaker 2>in classrooms. Like, the idea that technology has helped the

0:02:38.240 --> 0:02:42.840
<v Speaker 2>classroom is there, but it's also been that technology has

0:02:43.240 --> 0:02:45.560
<v Speaker 2>hurt the education experience. That's a big part of the

0:02:45.560 --> 0:02:50.440
<v Speaker 2>conversation right now. How do those two things conflict with

0:02:50.480 --> 0:02:53.280
<v Speaker 2>one another? How do you see them?

0:02:54.040 --> 0:02:57.240
<v Speaker 4>The reality is technology is always a it's neutral. It

0:02:57.240 --> 0:02:58.880
<v Speaker 4>can be positive or negative. You can think of a

0:02:58.960 --> 0:03:01.040
<v Speaker 4>very simple technology like a knife. It can be used

0:03:01.080 --> 0:03:02.520
<v Speaker 4>to hurt people, It could be used to cook, it

0:03:02.520 --> 0:03:04.400
<v Speaker 4>could be used to build things. And we're seeing the

0:03:04.480 --> 0:03:07.680
<v Speaker 4>same thing with cell phones and with AI. Yes, and

0:03:07.720 --> 0:03:09.440
<v Speaker 4>the cheating is a real issue and people are still

0:03:09.440 --> 0:03:11.120
<v Speaker 4>scrambling and you know, I have a whole chapter in

0:03:11.160 --> 0:03:13.240
<v Speaker 4>the book about cheating and the state of cheating. Even

0:03:13.280 --> 0:03:16.240
<v Speaker 4>before AI was actually pretty bad, AI just really put

0:03:16.280 --> 0:03:20.680
<v Speaker 4>a spotlight on it. Cell Phones, social media obviously addict students,

0:03:20.760 --> 0:03:22.960
<v Speaker 4>especially young people in a certain way, and most of

0:03:23.000 --> 0:03:25.919
<v Speaker 4>the literature it's not good for them. But even before AI,

0:03:25.960 --> 0:03:27.959
<v Speaker 4>there's really positive things you can do with technology. You

0:03:27.960 --> 0:03:30.080
<v Speaker 4>could learn on con academy. You can write a paper,

0:03:30.240 --> 0:03:34.200
<v Speaker 4>you can code, you could edit video, and most of us,

0:03:34.240 --> 0:03:35.760
<v Speaker 4>I think, as long as done in moderation, would be

0:03:35.760 --> 0:03:38.040
<v Speaker 4>happy if our students spend some time a few hours

0:03:38.080 --> 0:03:40.200
<v Speaker 4>a week doing that type of thing. And so the

0:03:40.280 --> 0:03:41.720
<v Speaker 4>key is how do you get how do you use

0:03:41.760 --> 0:03:44.200
<v Speaker 4>technology to amplify positive intent and then how do you

0:03:44.200 --> 0:03:46.960
<v Speaker 4>also use it to mitigate the negative intent? We're building

0:03:47.000 --> 0:03:48.600
<v Speaker 4>and I write about in the book where we're building

0:03:48.640 --> 0:03:51.120
<v Speaker 4>tools that not only support students better in their writing.

0:03:51.160 --> 0:03:53.360
<v Speaker 4>Won't do it for them, It'll tutor them, it'll coach them,

0:03:53.440 --> 0:03:55.720
<v Speaker 4>but it can make that process transparent for teachers. And

0:03:56.000 --> 0:03:59.320
<v Speaker 4>by doing that, it actually undermines cheating, whether it's AI

0:03:59.400 --> 0:03:59.960
<v Speaker 4>cheating or other.

0:04:00.840 --> 0:04:02.480
<v Speaker 1>How does it though, Like I think one of the

0:04:02.520 --> 0:04:06.040
<v Speaker 1>things the conversations we've been having TOIL is about kind

0:04:06.040 --> 0:04:08.240
<v Speaker 1>of this idea of does it take some of our

0:04:08.280 --> 0:04:09.440
<v Speaker 1>thinking away from us?

0:04:09.520 --> 0:04:11.800
<v Speaker 3>Like you know, sometimes you got to struggle.

0:04:11.400 --> 0:04:14.560
<v Speaker 1>Through a concept or a thought or a mathematical problem.

0:04:15.320 --> 0:04:17.480
<v Speaker 1>Getting an assist can be helpful, but sometimes you got

0:04:17.480 --> 0:04:18.839
<v Speaker 1>to sit with it for a little bit. And I

0:04:18.880 --> 0:04:22.960
<v Speaker 1>do wonder about technology taking away some of that work that.

0:04:23.880 --> 0:04:25.839
<v Speaker 3>We use our brain for. And I'm just a little

0:04:25.839 --> 0:04:26.600
<v Speaker 3>nervous about that.

0:04:28.040 --> 0:04:31.559
<v Speaker 4>It's a valid fear. And you know what I believe

0:04:31.600 --> 0:04:34.080
<v Speaker 4>is one in some cases you do want the shortcut

0:04:34.120 --> 0:04:36.799
<v Speaker 4>when you're working, et cetera. But in most cases, especially

0:04:36.800 --> 0:04:38.440
<v Speaker 4>for since we want them to develop those skills. And

0:04:38.440 --> 0:04:40.839
<v Speaker 4>that's why everything we've been working on con Migo, which

0:04:40.880 --> 0:04:44.400
<v Speaker 4>is our AI assistant on con Academy, it doesn't answer

0:04:44.480 --> 0:04:46.560
<v Speaker 4>the question. It will push you socratically. So the other

0:04:46.600 --> 0:04:48.480
<v Speaker 4>day I was trying to get it to explain supernova's

0:04:48.720 --> 0:04:50.279
<v Speaker 4>at a deeper level. I know what they are, but

0:04:50.400 --> 0:04:52.680
<v Speaker 4>exactly why does it explode? Why doesn't it just collapse?

0:04:53.120 --> 0:04:55.160
<v Speaker 4>And its response wasn't it just give me a Wikipedia

0:04:55.160 --> 0:04:57.000
<v Speaker 4>answer or like the type that you get on chat GPD.

0:04:57.200 --> 0:04:59.320
<v Speaker 4>It's response was, well, can you explain to me what

0:04:59.320 --> 0:05:02.320
<v Speaker 4>you already know? And then it kept pushing my thinking, Well, okay,

0:05:02.360 --> 0:05:03.640
<v Speaker 4>if that leads to this, well, what do you think

0:05:03.640 --> 0:05:05.640
<v Speaker 4>what happened there, and so it was a much more

0:05:05.680 --> 0:05:08.840
<v Speaker 4>socratic conversation that drives critical thinking. And so there's an

0:05:08.880 --> 0:05:11.800
<v Speaker 4>opportunity to actually scale that type of learning that's been

0:05:11.880 --> 0:05:12.679
<v Speaker 4>very hard in the past.

0:05:12.960 --> 0:05:17.320
<v Speaker 1>What what is your AI tool like? Where what data

0:05:17.320 --> 0:05:18.640
<v Speaker 1>are you pulling on for it?

0:05:20.600 --> 0:05:24.599
<v Speaker 4>So we are using the underlying frontier models, whether it's

0:05:24.640 --> 0:05:27.400
<v Speaker 4>you know now GBT four, Turbau and on me, et cetera.

0:05:27.880 --> 0:05:29.760
<v Speaker 4>But then on top of that, there's a lot of

0:05:29.800 --> 0:05:32.880
<v Speaker 4>work to make it really special purpose for education. We're

0:05:32.920 --> 0:05:34.520
<v Speaker 4>doing a lot of what's called in the trade now

0:05:34.560 --> 0:05:38.799
<v Speaker 4>prompt well, prompt engineering, prompt chaining, multiple calls. It's anchored

0:05:38.800 --> 0:05:42.240
<v Speaker 4>on con academy content that's been vetted by professionals for

0:05:42.520 --> 0:05:44.800
<v Speaker 4>many years. It can slurp that stuff in as well.

0:05:45.440 --> 0:05:47.760
<v Speaker 4>So we're putting a lot of the pieces together. We're

0:05:47.800 --> 0:05:50.240
<v Speaker 4>not just you know, relying on the raw model, which

0:05:50.240 --> 0:05:52.320
<v Speaker 4>I don't think is a responsible way to do it

0:05:52.320 --> 0:05:53.719
<v Speaker 4>for an education use case.

0:05:54.120 --> 0:05:57.760
<v Speaker 2>It's still it is a brave new world that we're

0:05:58.600 --> 0:06:03.160
<v Speaker 2>living in. Your book's called Brave New Words. Yesterday saw

0:06:03.160 --> 0:06:05.240
<v Speaker 2>I don't know if you saw this, but a lot

0:06:05.240 --> 0:06:10.120
<v Speaker 2>of people were posting on x about errors that Google's

0:06:10.160 --> 0:06:13.600
<v Speaker 2>own AI search tool was making. My brother sent me

0:06:13.640 --> 0:06:17.080
<v Speaker 2>one about eating rocks, and it actually had cited an

0:06:17.120 --> 0:06:20.480
<v Speaker 2>Onion article that was joking about eating rocks. But you know,

0:06:20.520 --> 0:06:23.400
<v Speaker 2>if you looked at eating rocks on the internet, it

0:06:23.440 --> 0:06:25.839
<v Speaker 2>said like it's recommended to eat rocks every day, and

0:06:25.839 --> 0:06:27.440
<v Speaker 2>then you look at the source and it's the oniony,

0:06:27.760 --> 0:06:29.920
<v Speaker 2>So like, this is not you know, this is by

0:06:29.960 --> 0:06:32.440
<v Speaker 2>no means like an easy thing to figure out at

0:06:32.440 --> 0:06:32.839
<v Speaker 2>this point.

0:06:32.839 --> 0:06:36.520
<v Speaker 4>So no, not not at all. And obviously I can't

0:06:37.000 --> 0:06:40.760
<v Speaker 4>I can't defend everything out there, but regardless, obviously the

0:06:40.760 --> 0:06:42.640
<v Speaker 4>providers are going to work harder to make sure that

0:06:42.680 --> 0:06:45.880
<v Speaker 4>things like that don't happen. But it's part of digital

0:06:45.880 --> 0:06:48.840
<v Speaker 4>literacy now for users to recognize that every answer you're

0:06:48.839 --> 0:06:50.320
<v Speaker 4>going to get from an AI, you're gonna have to

0:06:50.400 --> 0:06:52.239
<v Speaker 4>validate that, You're gonna have to put that in context,

0:06:52.240 --> 0:06:54.920
<v Speaker 4>and not all AIS are considered equal. And this isn't

0:06:54.920 --> 0:06:58.480
<v Speaker 4>a new phenomenon. Well before generative AI, you could do

0:06:58.520 --> 0:07:02.039
<v Speaker 4>a web search on Google and very click on something

0:07:02.080 --> 0:07:04.880
<v Speaker 4>that's complete misinformation, or see a video on YouTube that's

0:07:04.880 --> 0:07:08.240
<v Speaker 4>complete misinformation. So we've always had to try to give

0:07:08.279 --> 0:07:10.240
<v Speaker 4>people that scale of how do they judge, how can

0:07:10.280 --> 0:07:12.640
<v Speaker 4>they validate? I think that's true in an AI world,

0:07:12.640 --> 0:07:14.240
<v Speaker 4>and I read a lot about it in the book.

0:07:14.400 --> 0:07:16.640
<v Speaker 4>There are ways, and we are building these ways that

0:07:16.720 --> 0:07:19.520
<v Speaker 4>the AI can not only not do those things or

0:07:19.600 --> 0:07:21.840
<v Speaker 4>can help you validate what's real.

0:07:22.160 --> 0:07:25.080
<v Speaker 1>Listen, you are someone who was so at the frontier

0:07:25.200 --> 0:07:28.200
<v Speaker 1>in terms of figuring out a really smart way to

0:07:28.280 --> 0:07:31.600
<v Speaker 1>use technology to teach kids, and they still do it.

0:07:31.640 --> 0:07:34.160
<v Speaker 1>I mean you just think about what you did way

0:07:34.160 --> 0:07:36.320
<v Speaker 1>back when, right when it was kind of I hate

0:07:36.360 --> 0:07:37.679
<v Speaker 1>to say, like almost simpler times.

0:07:37.800 --> 0:07:40.600
<v Speaker 2>She doesn't hate YouTube, right, it's very early.

0:07:40.480 --> 0:07:43.240
<v Speaker 1>It's a while, but it has helped so many and

0:07:43.280 --> 0:07:44.840
<v Speaker 1>continues to do so.

0:07:44.840 --> 0:07:46.680
<v Speaker 3>So, so you like.

0:07:48.720 --> 0:07:51.200
<v Speaker 1>Not glass half full with AI and this, Like you

0:07:51.240 --> 0:07:53.400
<v Speaker 1>think it's going to be okay, You think it's going

0:07:53.440 --> 0:07:56.720
<v Speaker 1>to be leaps and bounds in terms of teaching folks

0:07:56.800 --> 0:07:59.440
<v Speaker 1>like I think about tutors. But anyway, like you are

0:07:59.520 --> 0:08:01.480
<v Speaker 1>all in and you feel comfortable about this.

0:08:02.560 --> 0:08:03.760
<v Speaker 4>Well, what I've told everyone. You know a lot of

0:08:03.800 --> 0:08:05.800
<v Speaker 4>people are reading the book. I'm getting interviewed together like, hey,

0:08:05.840 --> 0:08:07.520
<v Speaker 4>you tend to be more optimistic than most people. I

0:08:07.560 --> 0:08:09.320
<v Speaker 4>was like, no, look, I'm afraid of all the same

0:08:09.320 --> 0:08:11.760
<v Speaker 4>stuff that a lot of people are saying. Bad people

0:08:11.800 --> 0:08:13.560
<v Speaker 4>are going to use it to amplify bad intent. We're

0:08:13.600 --> 0:08:15.200
<v Speaker 4>going to see deep fakes, we're going to see fraud,

0:08:15.200 --> 0:08:17.320
<v Speaker 4>We're going to see state actors try to manipulate people

0:08:17.360 --> 0:08:19.840
<v Speaker 4>in bad ways. That's going to happen. But that doesn't

0:08:19.880 --> 0:08:22.240
<v Speaker 4>mean that the good people should just freeze and do nothing.

0:08:22.520 --> 0:08:25.720
<v Speaker 4>We should use these tools to amplify positive intent. And

0:08:25.760 --> 0:08:27.760
<v Speaker 4>so what I'm telling everyone, whether AI is a net

0:08:27.800 --> 0:08:30.160
<v Speaker 4>positive or net negative, isn't like a flip of the coin.

0:08:30.160 --> 0:08:32.280
<v Speaker 4>We're not just going to like sit and watch. Good

0:08:32.280 --> 0:08:34.640
<v Speaker 4>people should go out there and make the positive use

0:08:34.679 --> 0:08:37.840
<v Speaker 4>cases happen. We shouldn't just talk about it. And they exist,

0:08:37.880 --> 0:08:39.960
<v Speaker 4>and we are already seeing that in education, you know,

0:08:40.080 --> 0:08:42.680
<v Speaker 4>just earlier this week, we made a big announcement with Microsoft.

0:08:42.720 --> 0:08:44.600
<v Speaker 4>We're making these teacher tools that are going to save

0:08:44.600 --> 0:08:47.360
<v Speaker 4>teachers time doing things like lesson planning and grading papers

0:08:47.400 --> 0:08:50.199
<v Speaker 4>and writing progress supports, saving teachers hours a week. We're

0:08:50.200 --> 0:08:51.960
<v Speaker 4>making that free, even though it costs a lot of

0:08:51.960 --> 0:08:55.120
<v Speaker 4>money for the computation, making that free to all us teachers.

0:08:55.480 --> 0:08:57.400
<v Speaker 4>So there's a lot of good that's already happening, and

0:08:57.400 --> 0:09:00.520
<v Speaker 4>so I'm optimistic if good people put their energy and

0:09:00.520 --> 0:09:02.280
<v Speaker 4>the resources behind it, we're going to get a net

0:09:02.320 --> 0:09:03.640
<v Speaker 4>benefit from the technology.

0:09:03.760 --> 0:09:05.920
<v Speaker 2>As Carol mentioned, you harnessed tech early in your career

0:09:05.960 --> 0:09:09.520
<v Speaker 2>two thousand and eight for YouTube. Is AI, in your opinion,

0:09:09.520 --> 0:09:10.439
<v Speaker 2>more powerful than that?

0:09:11.559 --> 0:09:13.680
<v Speaker 4>Oh yeah, I think you know, even if we just

0:09:13.760 --> 0:09:16.520
<v Speaker 4>even if they froze AI today, there's ten years of

0:09:16.600 --> 0:09:19.040
<v Speaker 4>amazing things we could create with it. But the reality

0:09:19.120 --> 0:09:22.000
<v Speaker 4>is the pace of innovations only accelerating. So yes, what

0:09:22.040 --> 0:09:23.240
<v Speaker 4>we're going to be able to do in two, three,

0:09:23.240 --> 0:09:26.000
<v Speaker 4>four or five years is going to be mind blowing, unbelievable.

0:09:26.160 --> 0:09:28.360
<v Speaker 2>I want to get right back to Salcon, a familiar

0:09:28.440 --> 0:09:30.520
<v Speaker 2>name to so many people. He's the founder and CEO

0:09:30.600 --> 0:09:33.120
<v Speaker 2>of the nonprofit con Academy as well as the con

0:09:33.200 --> 0:09:34.960
<v Speaker 2>Lap School. He's got a new book out. It's called

0:09:35.000 --> 0:09:38.760
<v Speaker 2>Brave New Words, How AI will revolutionize education and why

0:09:39.000 --> 0:09:41.720
<v Speaker 2>that's a good thing. Sal We're going to get back

0:09:41.720 --> 0:09:43.000
<v Speaker 2>to the book in just a second. But as we

0:09:43.040 --> 0:09:45.040
<v Speaker 2>were on our planning call this morning, we were talking

0:09:45.120 --> 0:09:50.480
<v Speaker 2>about con Academy and we're all sharing our experiences with it,

0:09:50.520 --> 0:09:53.120
<v Speaker 2>and it feels like with you because those early videos

0:09:53.120 --> 0:09:57.280
<v Speaker 2>were really you and you you only teaching us so much,

0:09:58.880 --> 0:10:01.480
<v Speaker 2>and we're asking ourselves, is it a nonprofit still? And

0:10:01.559 --> 0:10:02.880
<v Speaker 2>you know, went up and looked, and it still is

0:10:02.920 --> 0:10:05.320
<v Speaker 2>a nonprofit. And I'm wondering talk to us a little

0:10:05.320 --> 0:10:08.760
<v Speaker 2>about the structure of the organization and why a nonprofit

0:10:08.800 --> 0:10:10.600
<v Speaker 2>is the right model for you.

0:10:12.200 --> 0:10:13.760
<v Speaker 4>Yeah, and it's always good to hear those stories. And

0:10:14.120 --> 0:10:17.320
<v Speaker 4>I still make a lot of videos. I made a

0:10:17.320 --> 0:10:18.160
<v Speaker 4>couple this morning.

0:10:18.240 --> 0:10:18.800
<v Speaker 2>That's amazing.

0:10:18.960 --> 0:10:23.040
<v Speaker 4>But you know, there was there's three Harvard Business Schools

0:10:23.120 --> 0:10:27.040
<v Speaker 4>case studies on con Academy, and the question that it

0:10:27.080 --> 0:10:29.560
<v Speaker 4>normally boils down to is should kind Academy be a

0:10:29.559 --> 0:10:31.960
<v Speaker 4>for profit or nonprofit? We clearly have a social mission,

0:10:32.000 --> 0:10:35.880
<v Speaker 4>but we have some of the behaviors or the properties

0:10:35.920 --> 0:10:38.720
<v Speaker 4>of a high growth for profit. We're very tech focused.

0:10:38.760 --> 0:10:41.120
<v Speaker 4>We're out here in Silicon Valley, we scale, et cetera,

0:10:41.160 --> 0:10:43.440
<v Speaker 4>et cetera. You know, for me, it was honestly an

0:10:43.480 --> 0:10:46.360
<v Speaker 4>emotional decision. When I incorporated con Academy back in two

0:10:46.360 --> 0:10:49.400
<v Speaker 4>thousand and eight. There were some vcs who were interested,

0:10:49.440 --> 0:10:51.959
<v Speaker 4>and I'm not against for profit. My previous life, I

0:10:52.000 --> 0:10:54.600
<v Speaker 4>was a hedge fund analyst, which was very for profit,

0:10:55.400 --> 0:10:59.280
<v Speaker 4>but it felt like education and it still feels that

0:10:59.280 --> 0:11:01.760
<v Speaker 4>way is a space where maybe the markets don't always

0:11:01.840 --> 0:11:04.480
<v Speaker 4>lead to outcomes that are optimal or are outcomes that

0:11:04.480 --> 0:11:06.960
<v Speaker 4>are aligned with our values. And there were folks even

0:11:07.000 --> 0:11:09.240
<v Speaker 4>back in two thousand and eight who were using the content,

0:11:09.280 --> 0:11:11.720
<v Speaker 4>who were telling me how it was changing their life,

0:11:11.760 --> 0:11:13.040
<v Speaker 4>and I was like, well, I'm not sure if they

0:11:13.040 --> 0:11:15.240
<v Speaker 4>would be able to access it if they didn't have it.

0:11:15.640 --> 0:11:17.560
<v Speaker 4>And as a hedge fun analyst, I also saw how

0:11:17.880 --> 0:11:20.160
<v Speaker 4>you could have a founder who really is mission driven

0:11:20.400 --> 0:11:22.520
<v Speaker 4>of a for profit, but you fast forward fifty years,

0:11:22.559 --> 0:11:26.280
<v Speaker 4>one hundred years, the capital structure changes, the incentives change,

0:11:26.400 --> 0:11:28.000
<v Speaker 4>and I didn't want that to happen. So I did

0:11:28.040 --> 0:11:30.000
<v Speaker 4>a kind of delusional thing. I was in this walk

0:11:30.000 --> 0:11:32.120
<v Speaker 4>in closet when I filled out the paperwork with the RS,

0:11:32.160 --> 0:11:34.400
<v Speaker 4>I said, what if Kin Academy could be like the

0:11:34.400 --> 0:11:37.320
<v Speaker 4>next Oxford or the next Smithsonian. Even at that time,

0:11:37.360 --> 0:11:39.600
<v Speaker 4>we had actually served more students than Harvard had in

0:11:39.600 --> 0:11:41.640
<v Speaker 4>its history, even when I was just one guy in

0:11:41.679 --> 0:11:43.360
<v Speaker 4>a walk in closet, And I said, what if we

0:11:43.360 --> 0:11:45.960
<v Speaker 4>could reach billions? And so it was a little bit

0:11:45.960 --> 0:11:49.120
<v Speaker 4>of a delusional optimism, but you know, it feels less

0:11:49.120 --> 0:11:51.960
<v Speaker 4>delusional now really interesting.

0:11:52.320 --> 0:11:53.719
<v Speaker 1>I do want to go back to the book, if

0:11:53.720 --> 0:11:57.199
<v Speaker 1>we may, your book Brave New Words, how AI will

0:11:57.280 --> 0:12:01.800
<v Speaker 1>revolutionize education and why that's a good thing. So where

0:12:01.800 --> 0:12:03.880
<v Speaker 1>do you think AI will have the biggest.

0:12:03.600 --> 0:12:07.560
<v Speaker 3>Impact on education? Is it the approach? Is it a

0:12:07.559 --> 0:12:11.160
<v Speaker 3>certain area? I'm just curious.

0:12:12.280 --> 0:12:16.080
<v Speaker 4>You know, the technology is so powerful. I think every

0:12:16.120 --> 0:12:18.600
<v Speaker 4>hour that you spend brainstorming, you realize it'll have more

0:12:18.600 --> 0:12:22.280
<v Speaker 4>and more implications. I mean in any industry, especially education.

0:12:23.400 --> 0:12:26.040
<v Speaker 4>You know when open Ai first reached out to us

0:12:26.040 --> 0:12:28.000
<v Speaker 4>in summer of twenty twenty two, and this was six

0:12:28.040 --> 0:12:30.880
<v Speaker 4>months before Chat GPT came out and we were able

0:12:30.880 --> 0:12:33.720
<v Speaker 4>to get early access to GPT four. Many folks know

0:12:33.800 --> 0:12:36.679
<v Speaker 4>that even the first chat GPT wasn't built on GPT four,

0:12:36.800 --> 0:12:41.360
<v Speaker 4>was built on a previous model. We immediately gravitated to, hey,

0:12:41.400 --> 0:12:43.480
<v Speaker 4>how can this be a tutor for any student? Really

0:12:43.520 --> 0:12:45.120
<v Speaker 4>do the types of things that I was doing with

0:12:45.160 --> 0:12:48.400
<v Speaker 4>my cousins twenty years ago and these frontier models now

0:12:48.800 --> 0:12:52.520
<v Speaker 4>like GPT four, Gemini, et cetera, they really can emulate

0:12:52.559 --> 0:12:55.920
<v Speaker 4>what a tutor could do via chat. And many folks

0:12:55.960 --> 0:12:59.439
<v Speaker 4>have probably seen the demo that for Omni GPT four

0:12:59.480 --> 0:13:02.240
<v Speaker 4>Omni that came out last week. It's me and my son.

0:13:02.679 --> 0:13:04.040
<v Speaker 4>They brought us in to see if we can do

0:13:04.080 --> 0:13:06.560
<v Speaker 4>some fun stuff with it, and you can talk to

0:13:06.600 --> 0:13:08.360
<v Speaker 4>it in a natural way, and it could see your

0:13:08.400 --> 0:13:11.200
<v Speaker 4>writing and it can see your face, and it really

0:13:11.200 --> 0:13:15.480
<v Speaker 4>felt like a live tutoring session. That's all coming. But

0:13:15.520 --> 0:13:17.160
<v Speaker 4>then there's the other side of the equation. There's what

0:13:17.160 --> 0:13:19.560
<v Speaker 4>you could do with teachers, helping them write lesson plans,

0:13:20.000 --> 0:13:22.600
<v Speaker 4>helping be a co teacher for that teacher, be a

0:13:22.600 --> 0:13:25.880
<v Speaker 4>teaching assistant, helping them grade papers. That's already happening. We

0:13:25.960 --> 0:13:29.360
<v Speaker 4>already getting feedback from school districts using Conmego that their

0:13:29.400 --> 0:13:31.480
<v Speaker 4>teachers are saving five to ten hours a week. That's

0:13:31.520 --> 0:13:34.440
<v Speaker 4>a huge benefit for teachers who are spread thin or

0:13:34.480 --> 0:13:37.320
<v Speaker 4>in many cases overworked. So it's a benefit for them.

0:13:37.640 --> 0:13:40.680
<v Speaker 4>But I see us going five years into the future

0:13:40.720 --> 0:13:44.840
<v Speaker 4>where it's even able to facilitate interactions between people. A

0:13:44.880 --> 0:13:47.720
<v Speaker 4>teacher can say, hey, let's do a breakout. Hey, I says, okay,

0:13:48.080 --> 0:13:50.280
<v Speaker 4>I'll set up the breakout rooms, and actually I will

0:13:50.280 --> 0:13:53.120
<v Speaker 4>moderate the conversation in each breakout and I will bring

0:13:53.160 --> 0:13:55.440
<v Speaker 4>it back to you. An AI and this we have

0:13:55.480 --> 0:13:57.720
<v Speaker 4>to just be careful that the data privacy and it

0:13:57.720 --> 0:13:59.840
<v Speaker 4>doesn't get creepy. But I can say, hey, we haven't

0:13:59.840 --> 0:14:02.760
<v Speaker 4>called Mary in a little while, why don't we call

0:14:02.800 --> 0:14:06.080
<v Speaker 4>on her. She looks disengaged or he looks disengaged. Let's

0:14:06.160 --> 0:14:08.679
<v Speaker 4>use that to personalize the lessons more so it really

0:14:08.720 --> 0:14:10.040
<v Speaker 4>will feel like a teaching assistant.

0:14:10.120 --> 0:14:14.199
<v Speaker 2>Heysel, you mentioned con migo and conmego is featured prominently

0:14:14.440 --> 0:14:16.719
<v Speaker 2>in your book. Can you explain what it is and

0:14:16.760 --> 0:14:18.960
<v Speaker 2>how you guys built it on GPT for.

0:14:20.520 --> 0:14:24.120
<v Speaker 4>Yeah, well, this is back in almost almost two years ago,

0:14:24.160 --> 0:14:26.960
<v Speaker 4>eighteen months ago. We saw there's a lot of potential

0:14:27.120 --> 0:14:28.920
<v Speaker 4>for the raw technology, but we also said, hey, this

0:14:28.960 --> 0:14:31.680
<v Speaker 4>could be used for cheating. If you're talking about kids,

0:14:31.800 --> 0:14:33.000
<v Speaker 4>how do you know that they're not trying to do

0:14:33.040 --> 0:14:34.880
<v Speaker 4>shady things with the AI? How do you protect their

0:14:34.920 --> 0:14:37.240
<v Speaker 4>data privacy? How do you make sure that it doesn't

0:14:37.240 --> 0:14:39.760
<v Speaker 4>make errors in math and hallucinate? And so we said

0:14:39.760 --> 0:14:41.840
<v Speaker 4>there's a lot of work here to take this raw

0:14:41.960 --> 0:14:45.800
<v Speaker 4>AI model and make it appropriate for school. So our

0:14:46.160 --> 0:14:49.400
<v Speaker 4>AI tool set for teachers and students we call con Migo,

0:14:49.960 --> 0:14:52.560
<v Speaker 4>and it does all of That's what it's doing. It

0:14:52.600 --> 0:14:55.280
<v Speaker 4>won't cheat, it will work with you socratically. It'll actually

0:14:55.280 --> 0:14:58.800
<v Speaker 4>help undermine other forms of cheating, even AI based cheating.

0:14:59.040 --> 0:15:02.680
<v Speaker 4>It makes everything to the teacher. It'll actively notify parents

0:15:02.680 --> 0:15:05.120
<v Speaker 4>and teachers if students try to do something shady with

0:15:05.160 --> 0:15:08.720
<v Speaker 4>the AI. We're introducing interfaces. Not everything should be a

0:15:08.800 --> 0:15:11.080
<v Speaker 4>chat interface. We're doing. The interface is where a student

0:15:11.080 --> 0:15:14.760
<v Speaker 4>can write an essay, co create a document. Once again,

0:15:14.800 --> 0:15:17.160
<v Speaker 4>all of that is transparent to the teacher. The AI

0:15:17.240 --> 0:15:20.240
<v Speaker 4>can highlight parts of the document and give students feedback

0:15:20.280 --> 0:15:22.800
<v Speaker 4>on it. We're created activities where it can act as

0:15:23.080 --> 0:15:26.600
<v Speaker 4>a bit of a career and guidance counselor for a teacher,

0:15:26.880 --> 0:15:29.440
<v Speaker 4>for a student. Once again, not to replace the human

0:15:29.480 --> 0:15:32.040
<v Speaker 4>beings in their life, but really to amplify that impact.

0:15:32.160 --> 0:15:33.960
<v Speaker 4>Whenever student needs it, it's there, and then when they

0:15:33.960 --> 0:15:36.200
<v Speaker 4>get with the teacher or their guidance counselor, they can

0:15:36.240 --> 0:15:36.720
<v Speaker 4>go deeper.

0:15:36.920 --> 0:15:39.000
<v Speaker 1>Hey, listen, one thing we talk about a lot salads,

0:15:39.040 --> 0:15:41.640
<v Speaker 1>you would imagine is college admissions. And you have a

0:15:41.720 --> 0:15:43.920
<v Speaker 1>chapter towards the end of the book devoted to that.

0:15:44.040 --> 0:15:46.760
<v Speaker 1>The first line in it says, the classic components considered

0:15:46.760 --> 0:15:51.400
<v Speaker 1>in college admissions are grades, standardized tests, extracurriculars, essays, and

0:15:51.480 --> 0:15:55.360
<v Speaker 1>letters of recommendation. AI will change how most, if not all,

0:15:55.440 --> 0:15:58.480
<v Speaker 1>of these factors are value developed and evaluated. Those are

0:15:58.520 --> 0:16:01.240
<v Speaker 1>your words, What impact will that have on that process?

0:16:02.800 --> 0:16:05.120
<v Speaker 4>It is really across the board. As we know, these

0:16:05.120 --> 0:16:07.920
<v Speaker 4>things are all about writing. And what we do also

0:16:07.960 --> 0:16:09.840
<v Speaker 4>know is, you know, there was a Varsity Blue scandal

0:16:09.880 --> 0:16:15.440
<v Speaker 4>several years ago where we saw these corrupt coaches, admissions

0:16:15.440 --> 0:16:19.360
<v Speaker 4>coaches that you could hire to almost fabricate a college admissions.

0:16:19.400 --> 0:16:21.880
<v Speaker 4>But what we know, you know, and especially places like

0:16:21.920 --> 0:16:24.240
<v Speaker 4>Silicon Valley where I live, New York, a lot of

0:16:24.280 --> 0:16:28.000
<v Speaker 4>affluent people, upper middle class people. Well before AI they

0:16:28.000 --> 0:16:30.520
<v Speaker 4>were paying four hundred dollars they are it's still happening,

0:16:30.520 --> 0:16:32.800
<v Speaker 4>paying four or five hundred dollars an hour to have

0:16:32.880 --> 0:16:37.200
<v Speaker 4>these professionals advise their kids on how to maximize their essays,

0:16:37.200 --> 0:16:39.600
<v Speaker 4>et cetera. Now, the ethical ones are just act as coaches,

0:16:39.640 --> 0:16:41.760
<v Speaker 4>but it's still a huge advantage, and the less ethical

0:16:41.760 --> 0:16:46.840
<v Speaker 4>ones are essentially writing these essays. And you know, I

0:16:46.840 --> 0:16:49.360
<v Speaker 4>think admissions officers, and I write about this in the book,

0:16:49.640 --> 0:16:52.240
<v Speaker 4>have they've kind of known about this, but they've kind

0:16:52.320 --> 0:16:56.920
<v Speaker 4>of intentionally not wanted to address this, the elephant in

0:16:56.960 --> 0:16:59.080
<v Speaker 4>the room about the inequity and to the fact that

0:16:59.120 --> 0:17:02.360
<v Speaker 4>the AIS can now do some of this work for

0:17:02.440 --> 0:17:04.960
<v Speaker 4>all students. It really just puts a spotlight on something

0:17:05.000 --> 0:17:08.439
<v Speaker 4>that's always there and it makes you wonder whether you

0:17:08.480 --> 0:17:13.640
<v Speaker 4>can get richer points of data for a student than

0:17:13.680 --> 0:17:17.720
<v Speaker 4>a recommendation than an essay. But that's where the AI

0:17:17.760 --> 0:17:22.920
<v Speaker 4>could be interesting too. Future admissions SAT type things could

0:17:22.960 --> 0:17:25.399
<v Speaker 4>have free response. You could have to talk things through,

0:17:25.920 --> 0:17:27.600
<v Speaker 4>you could write it out and then the AI could

0:17:27.600 --> 0:17:28.240
<v Speaker 4>make sense of that.

0:17:28.400 --> 0:17:28.840
<v Speaker 3>So cool.

0:17:28.880 --> 0:17:31.399
<v Speaker 1>There's also a book about a chapter about employment in

0:17:31.440 --> 0:17:32.120
<v Speaker 1>an AI world.

0:17:32.160 --> 0:17:33.399
<v Speaker 3>There's so much in this book.

0:17:33.480 --> 0:17:36.400
<v Speaker 1>Sal thank you for finding time for us to talk

0:17:36.440 --> 0:17:40.040
<v Speaker 1>about sol Con. Of course, his new book, Brave New Words,