a16z Podcast - Classroom 2050: Unleashing AI, XR, Gaming

Episode Date: August 14, 2023

With students learning in more places and different ways than we have ever seen, the pace of change in education is dizzying. Can this progress narrow learning gaps exacerbated by COVID-19 or will it fuel divides? How can we make the most of this once-in-a-generation opportunity?Join our panelists from Khan Academy, PrismsVR, and Minecraft Education as we discuss the classroom of the future, and how important technologies including AI, XR, and gaming should play a role in it. Topics Covered:00:00 – The classroom of 202304:30 – Data on the current classroom05:40– Developing the next gen of STEM talent08:00 –  AI in the classroom 13:10 – The role of gaming in the classroom16:00 – Misconceptions of technology in classroom17:51 – The ways teachers and students use AI21:15 - The future role of a teacher24:25 – Personalization and ownership in education27:11 – Getting hardware into the classroom29:54 - Misconceptions of teacher’s abilities and appetite for tech31:25 – Why implement these changes now?36:05 – Call to action for the classroom of 2050 Resources:Find Sal Khan on Twitter: https://twitter.com/salkhanacademyFind Romy Drucker on Twitter: https://twitter.com/romydruckerFind Allison Matthews on LinkedIn: https://www.linkedin.com/in/allison-matthews-4050677/Find Anurupa on LinkedIn: https://www.linkedin.com/in/anurupa-ganguly-92790379/Learn more about Khan Academy: https://www.khanacademy.org/Learn more about The Walton Family Foundation: https://www.waltonfamilyfoundation.org/Learn more about Minecraft Education: https://education.minecraft.net/en-usLearn more about Prisms VR: https://www.prismsvr.com/Watch more panel discussions from The Aspen Ideas Festival 2023: https://www.aspenideas.org/ Stay Updated: Find a16z on Twitter: https://twitter.com/a16zFind a16z on LinkedIn: https://www.linkedin.com/company/a16zSubscribe on your favorite podcast app: https://a16z.simplecast.com/Follow our host: https://twitter.com/stephsmithioPlease note that the content here is for informational purposes only; should NOT be taken as legal, business, tax, or investment advice or be used to evaluate any investment or security; and is not directed at any investors or potential investors in any a16z fund. a16z and its affiliates may maintain investments in the companies discussed. For more details please see a16z.com/disclosures. 

The situation is not good, and a lot of people blame the pandemic, but it was pretty bad before the pandemic. The kids in school today are growing up into a world that's incredibly fast-paced, and it's complex. I think a common misconception is that technology is going to replace the pedagogy, and it's not. Many parents are thinking differently about what learning looks like, learning on a continuum, different experiences. In Minecraft, they're like figuring out who they are as a human. Like if they're the person playing with their friends who's organizing the mission, maybe they're going to grow up to be a leader. It is not a clear path to Classroom 2050.

So we have to be able to hold the optimism and hope and boldness. It's like the holodeck on Star Trek, and I honestly didn't think that was going to happen in my lifetime. I think that's going to happen in the next five years. I know this panel was called the Classroom of 2050, but I actually want to be very clear. It's a classroom of 2023. It's happening right now. The year is 2023, and so much has changed in the last 50 years when it comes to technology. We now have smartphones that can fit in our pockets, we have the world's knowledge accessible to anyone with an internet connection, and even tools that summarize incredibly complex concepts in the voice of our favorite fictional characters.

Now, reflect on the state of the classroom, the very forum where we raise the next generation. How is that evolved during the same time period? Despite so many children actively using cutting-edge technologies like AI, XR, and gaming outside the classroom, those technologies haven't made their way in. And this episode explores why that is, especially given the learning loss we saw prior to and during COVID. and how these technologies just might help us radically upgrade the classroom of the future. This episode was actually recorded live at the Aspen Ideas Festival, which I got to attend for the first time recently,

and I was joined by four guests with such unique backgrounds. The first was Sal Khan, founder of Khan Academy, which recently unveiled Khan Migo, an AI-powered tutor. We also had Anorupa Ganguli. She is the founder of Prisms VR, which is using VR to better teach children math and science through spatial reasoning. We also had Alison Matthews, head of Minecraft education, who's tailoring the world's most popular game to empower and teach young people skills from coding to active citizenship.

And finally, we had Romi Drucker, director of the education program at the Walton Family Foundation, who has a unique bird's-eye view of the challenges and opportunities in the space, also previously serving in leadership roles at the New York City Department of Education. I really wish I grew up in an era where these technologies were at my fingertips. So let's not squander this unique opportunity. As a reminder, the content here is for informational purposes only. Should not be taken as legal, business, tax, or investment advice, or be used to evaluate any investment or security,

and is not directed at any investors or potential investors in any, A16Z fund. Please note that A16Z and its affiliates may also maintain investments in the companies discussed in this podcast. For more details, including a link to our investments, please see A16C.com slash disposures. Welcome, everyone. Welcome to Classroom 2050. Welcome to our classroom for the next 50 minutes. I'd love to just start by getting a pulse from the room. How many of you are? are excited about new technologies like AI, XR, extended reality, gaming. Let's put your hands up.

Who's excited? Great. Now, how many of you, keep your hands up and then put them down, how many of you feel like these technologies have successfully made it into the classroom for the better? Put your hand down if you don't think they've made it there. Keep it up. I only see a few hands still up there. But luckily, we have some wonderful founders today who are working on exactly that problem,

figuring out how we elevate the classroom after it's been relatively stagnant, quite frankly, for many years. All right, so this is going to be an amazing panel. And to kick things off, let's really ground ourselves on where we are. Romi, maybe you can help us here. What is the data saying in terms of the student experience? What are kids really facing? How can we really get a better understanding of that?

Thank you so much, Steph, for being here and moderating this discussion. As many of you know, the pandemic was a huge setback in a lot. education and it exacerbated and widened already deep inequities. We know from data that came out last week that math achievement levels are the lowest they've been in this country since 1990. And literacy achievement levels are the lowest they've been since 2004. And the pandemic erased two decades of progress that we had made in education through a variety of different change initiatives. So we have a lot of work to do. The good news is that students, families, and educators agree that this is a moment for reimagining learning, both embracing technology, but also thinking

about how we strengthen relationships in classrooms, how we better support student well-being to address the mental health crisis. Thank you, Romi. And Arupa, I want to move to you. What will it take specifically to develop the next generation of STEM talent. I know you have some data there. Give us a grounding there in terms of what that will really take and what we're seeing up until now. Absolutely. I discovered that the top indicators of success in post-secondary STEM is number one, your ability to think spatially. The Soviets at first discovered it's in 1970s. So if you kind of think about a rough definition, it's your ability to rotate 3D objects in your mind and maintain perspective of that object. The second top indicator of success in post-secondary

is your ability to abstract. And when I say abstract, I don't mean looking at word problems, talking, drawing diagrams, and visualizations, creating charts and tables, and creating equations, because those are all too-dea representations of thought. Abstraction is Isaac Newton. He was sitting under a tree and Apple Fowle. He's like, whoa, how do I ascribe language, notation, and build mathematical models of this physical experience? So I started Prisms three years ago to scale these best practice methodologies. How do you immerse children in problems in the first person, create the intellectual need for why they have to learn something, and then through building a solution to that problem, learn the math and science that you need to.

So how does prisons work? We utilize virtual reality. The students put their VR headsets on, their mobile VR headsets, and they're transported to a, it is my mission and commitment that when kids see math problems in math class, they should not be like, that's not a real problem, no one really important to actually solving it. They should go, I heard that in the news. I know absolutely that very smart people are working on this, building mathematical identities from a very young age and is believed that I can contribute to the mathematical sciences.

So that through line of going from physicalities, simulations, data visualizations, charts, tables, equations is really scalable. It's most of secondary math and science. So we've built out all of algebra one geometry, middle school math, algebra two precalculus. We'll be pushing into all of higher ed very soon. When we put out math, we've got teachers, we're clamoring. So we put out all of biology, chemistry, physics, middle school life sciences. And we're going to keep building.

This has really been built on the back. of teachers and in collaboration in school systems. So I'm happy to kind of share more. We have so much to dig into there. And by the way, we didn't have a chance to call this out, but there's many ways that people learn. And instead of talking about the technologies that these wonderful founders are building,

we wanted to make sure you had a glimpse into really what we're talking about here. And so we've talked about different technologies. So this is a glimpse into VR. One of the other technologies entering the classroom is AI. One of the topics of the year. So Sal, why don't you talk to us about what you're building there? and, of course, we'll also pull up some of the technology that you're building.

Yeah, so Khan Academy's been around a little while now, and most folks associate us with videos, but most of our resources have been around personalized practice. But I'm kind of a traditionalist and a progressive, where I say, well, but there's also just gaps in students' foundations. The number of times that I've gone to like a seventh grade math classroom, this was recent. I was in the Bronx. They were working on Khan Academy. The kids were doing exponents, and it was like summing to the third to the seventh power. So it's to the 21st power, three times seven.

I saw an eighth grader grab a calculator for that. I literally slapped their hand. I don't know if you're allowed to do that. But I slapped their hand. I'm like, what's three times seven? And I saw the eighth graders start to do this. And that's just a very basic example of like there's massive fluency gaps in students. And it's causing too much cognitive load.

So that's where Khan Academy started is, hey, can we get people as much practice? Learn at their own time and pace. You're a teacher with 30 kids, but there is a way that all 30 kids can go at different paces. You at a teacher can understand where the kids are. And the students have multiple shots at goal in order to understand those underlying skills. That's what we've been working on for 10 plus years. And I'm happy to talk more about it, but we're seeing some very promising results, pre-AI. But then last summer, open AI reaches out to us. This is obviously months before chat GPT and all of that. And they said, hey, we're training our

next generation model was GPT4. When chat GPT came out, it was based on GPT 3.5. I was skeptical because I had been paying attention to previous generations of these models. But when we started playing with GPT4, we said this one's different. This one, you can actually get it to act as a tutor. It still had issues around hallucinations, which is when the AI just makes up facts. It still has, that is the technical term for it. It had issues when it came to math, which obviously, you know, we don't only do math, but math is very close to our heart, and it's got to be accurate. And we were worried about guardrails, and what if students use it for negative things or they go into some type of weird rabbit hole? So we said, oh, can we put guardrails around it?

So we started working pretty aggressively. We were under NDA with OpenAI since last summer. Then Chad GPT came out, and we weren't going to launch our thing until the GPT4 launch in March. And I remember telling Greg Brockman like, hey, I thought we were waiting. It was like, no, Chad GPT is nothing new. It's just a chat interface on top of GPT 3.5. And that captured everyone's imagination. And it kind of immediately put a bit of a bomb into the education system, where everyone was like,

oh, kids are going to use this to cheat, write their essays. by the way, it's bad at math, and it makes up things, et cetera, et cetera. I was initially bummed because, like, we're working on what I think is a reasonably thoughtful solution here. Now, this thing just gets thrown out into the universe, and everyone's first impression is going to be AI plus education is really bad. In hindsight, I actually think that was a blessing because what it did is it forced people to grapple with some real things.

When we first saw GPT4, months before chat GPT, we were like, oh, yeah, people are going to be able to use this to write essays, people are going to be able to use this to do colleges, you know, do their homework, et cetera. how do we mitigate these risks? But the fact that chat GPT, within a month or two, 40% of teachers were using this very imperfect tool for themselves for things like lesson planning. And students were already using it,

but it was forcing the issue on the education. Most of ed tech is things are okay. Maybe we could use technology to make it better, but this was like, wow, things just got broken. We need to do something to make it better. So we launched our implementation, which we call ConMigo, on Khan Academy. There's two headlines.

here is one is we think this is the beginning of a journey. We've always tried to approximate personal tutoring at Khan Academy through videos and exercises, but I think Generative AI kind of jumped us forward 20 years faster than I thought, if you had asked me a year ago, where it can act as a tutor, and maybe just as interestingly act as a teaching assistant to help generate lesson plans, help refresh a teacher's knowledge, create rubrics, etc., etc. These are real kids in Newark, New Jersey, who are using Conmigo. And in that last screen you just saw, it was actually in Spanish as well. It works in every major language, which by itself already solves a major issue,

where you can be doing your work in English, but get help in Brazilian Portuguese, in Spanish, in Vietnamese, whatever language you need that help in. It does not cheat if you ask conmigo, hey, tell me the answer. It will not tell you the answer, but say, I'm your tutor. It'll act Socratic and say, well, what do you think is the next step? We've done a bunch, unless you all want to, I won't go in all the details, but it's substantially better. I won't say perfect,

but substantially better at math than what you would get just out of the box with GPT4. All conversations are monitorable by parents and teachers, which we thought was an important guardrail. And then we have a second AI that monitors the conversations.

And if it looks like it's going into a weird place, self-harm, hate speech, whatever, the AI will actively notify the parents and teachers to do that. And so we think this is the beginning. There's a lot coming. We're giving it memories so it can have a long-lasting interactions with the teachers and students,

ways that it reports back to the teachers so it can summarize things. We're adding speech. Both founders so far have talked about maybe how we're early on. There are misconceptions, misperceptions from whether it's students, teachers, people in school boards about welcoming this technology. So I want to get into those conversations too. But Allison, please give us a glimpse into the world of gaming because that is a form of technology that has.

has immense growth as well, just like AI and extended reality, but maybe isn't met with the same level of open arms. So we'd love to hear what you're building. Yeah, for sure. Minecraft is an immersive 3D virtual space that's filled with beautiful biomes and all kinds of animals. So Minecraft education. We built on top of that wildly popular video game

a powerful teaching and learning platform by putting safety, security, privacy features that educators need and parents expect on top, as well as an in-game library that's full of learning experiences from learning to code and math, history, literacy, but all the way to learning how to be a more active citizen or learning about climate change. So why would we do this? because we know, and maybe you do too, that video games are one of the most powerful ways to engage a person in a deeply engaging, deeply immersive experience.

And it lights up intrinsic motivation. Intrinsic motivation is just that desire to keep trying to go deeper, to learn more because you love it. You're having fun. You're passionate about it. So I'll give you an example. I was in the kitchen. Maybe you can relate to this a little bit.

it's in the kitchen with my two teenage boys. And I just dropped a fact that I had learned, I had memorized in school about the moon phases. My boys both looked at me deeply confused. And in a few-minute conversation that involved plates and a flashlight, they helped me understand that this fact that I had memorized in school was completely wrong. And so how did they understand this more deeply than me? Because certainly I didn't teach them. It's because they were playing a video game where they needed to build a rocket that had to be able to orbit the moon

and in order for the rocket to orbit the moon they had to understand space science and so they were motivated to learn about space science and all of the factors that are involved in it because they wanted to win this game. And so that is the power of game-based learning. Amazing.

And I'm going to go to you next with these misconceptions partially because Aniripa is actually a previous educator herself. And so tell us a little bit more about maybe the way that people view the introduction of these technologies, whether it be VR, but also AI, also gaming. And where maybe some of those assumptions may be misplaced. Yeah, I think a common misconception is that technology is going to replace the pedagogy, and it's not. Pedagogy is pedagogy.

And so what the tech will allow you to do is it will scale that. You are building an elementary school building in India that creates a shaded region. and you're building a satellite that does XYZ to do that thing, you have to learn ratios. You have to apply linear functions. And so what I've learned is that instead of talking about the technology, talk about the learning design. And when you're focused on the learning design,

whether it's a powerful AGI tool or game design or in my case, spatial computing, that is just fueling a way of teaching and learning. And when you projected in that manner, nobody pushes back. Like I have 100% of teachers opting in to our program. And that might seem like, are you sure on Rupa? Like everyone's a little bit afraid of VR and no, because in my deck, you don't see VR anywhere. So that's the first thing that I would say is a key misconception is it's not about the technology.

The technology powers the learning design. And the second thing that I would say was it is then therefore incumbent upon those who are building learning solutions to over index on training teachers in the pedagogy and not the tech. We focus too much time on let me teach you about like how to utilize this hardware. No, we're going to spend 30 seconds on that because you're actually very smart. and you know how to do it. You just need to kind of have that confidence. And we're going to spend the market share of the time on lesson study, intellectual preparation, job-embedded coaching, how do you get learning outcomes? So I would say kind of that's where my mind to in terms of the main misconception that the bow that I would tie on that is there's been this feeling

and perception that school systems are regressive and there's somehow contra to progress. When you show them how it's standards aligned, how you teach bell to bell in a 50-minute instructional episode, how you prepare for it, you're not going to get any pushback. It's just kind of showing them how you go from point A to point B. Sal, maybe you can comment on that. You already touched on some ideas where, you know, there is this assumption that with AI in the classroom, you know, every student is going to use it to cheat.

You're never going to learn again because why would they, right? They've got this assistant who can do it for them. So maybe just speak to that directly. What are you seeing in terms of maybe the reality or what can be designed to maybe construe that reality? Yeah, I think Chad GPT has forced folks into one of two modes, either have some assignments where you just say, use whatever help you want.

And that's reasonable because we now use it in the workplace, et cetera. There's another mode, even without technology, everyone said, okay, I guess we're going to have to do more in-class proctoring now, things like that. I think both of those are probably part of the mix. One of the things we're trying to do, we think actually the most power happens when the AI works with the student. So in a couple of months, we're going to be launching something. We're going to initially do it for college essays, which is a whole other. This is the first college admission cycle in the chat GPT era.

That's true. It's going to be very interesting. But we've created activities where it can interview the student and help surface things that would make for good college essays. So brainstorm with the student, not for the student, a modality where you can write an essay of any kind and then ask for feedback from the AI. You know, we've all seen technologies that can help with your grammar and things like that. But once again, not just tell you this is how to write it grammatically correct, but say, hey, look at your verb tense again. Like once again, act like a tutor. But now with generative AI, especially with GPT4, it can really go into the argument that you're

and how do you make it more emotive, and how do you really make it an excellent paper? And what's interesting about that is when you work with the AI and all of the interactions are logged, a teacher can actually not just introspect on the final outcome, they can introspect on the process. So we're building a mechanism right now where the teacher can assign through the AI, hey, I want everyone in my 11th grade AP American History course to write a paper on, you know, how the world would have been different if the American Revolution didn't happen to a word. I don't know, something. And then every student will be able to work on it with Conmigo.

And Conmigo will be like, okay, do you think that really backs up your argument? How about this? Like, they're riffing together, which I think will be a better experience for a lot of students. And then even while that's happening, the teacher will be able to talk to Kanmigo. It's like, okay, what's everyone up to and says, okay, I've already started a, you know, this student is doing great. This student seems kind of stuck. This student just copied and pasted something in there.

I don't know where it came from. We might want to double click on that. Maybe they're using chat GPT. And I think then all of a sudden you've kind of addressed this issue where now the artificial intelligence not only do not have to worry about policing it in the same way

can help students write better it can give teachers better feedback a lot of mastery learning is multiple shots at goal if you're at a certain level 80% you should try again you should be able to writing but writing it's very hard to get feedback

but now the AI can give multiple rounds and then also give the teacher first pass of like you know based on the rubric we created together I'd give the student a four out of five on that dimension a five out of five on that dimension. So I think that's probably where we're going to go. By the way, this isn't crazy.

I did a year of exchange in Sweden, and they have a totally different grading system, including you can submit an assignment and you can pass or fail. And if you fail, they tell you what to change. They tell you how to get better. So this is not like completely, you know, grown out of AI and totally new.

These concepts do exist elsewhere. We've been doing that in our lab school with humans, but it's more resource intensive. Exactly. Allison, actually, that dovetails perfectly into this idea that if we are introducing these technologies, what is the future of a teacher or teachers in general? How does that role evolve?

And then also, if you have time, I'd love to just hear your take on how people view gaming because that's a very interesting topic on its own. Yeah, so Minecraft education is now in 119 countries, and we've got millions of players every month, and a big focus of the team is on connecting with educators, so we can understand what are the problems that they're trying to solve. and one of them is similar to what Sal was talking about

in that in any given classroom, you're going to have a full range of readiness for whatever the course material is that's being covered. And so what teachers really need is a tool that enables differentiated learning. So a way that someone who totally understands the content can take it further, and someone who's new to them,

they haven't understood this before, can start to get familiar with the materials. And to your question of the role of the teacher, Because the game supports, we've got scaffolded learning experiences in the game, which means if you're stuck, if you don't know Minecraft or if you don't really understand the lesson, we've got little things built into the lessons that can help give you tips or clues on how to move forward. So students can be self-directed in that way, which frees the teacher up to actually work with the students who need more attention or to challenge the students who are racing ahead. So that's one. I wanted to point out one other thing for teachers, because we've had a lot of teachers, especially early on, tell us the kids love this game. Why would I bring this into the classroom? I'm going to lose control immediately. They all understand it better than I do. So yes, this is true. And none of us as adults will ever understand Minecraft as well as the average eight-year-old. This is just a fact. And so what that requires is a mindset shift, right? The kids, in school today are growing up into a world that's incredibly fast-paced and it's complex. So they need mindset that's going to enable them to be successful in a world that's changing rapidly. So they

need to build a growth mindset. They need to build leadership skills, collaboration, communication skills. And these are sort of those higher order things that can be achieved through playing a game and by students being immersed in this game, stoked to be there, having so much fun, the teacher then has the ability, instead of trying to keep students on task, the teacher has the ability to generate some of these conversations in the classroom about.

Tell me why you did that in the game. Help me understand your thought process that made you go destroy that other students work in the game. Let's have a conversation about digital citizenship and the fact that, you know, What we do in a virtual world creates feelings in the real world. And this is something that, you know, it's really important for all young people and adults, frankly,

to learn, like, how can we be better citizens in the world? So just sort of an example of taking things to the next level. I love that. Romie, how can the many folks from different backgrounds who have different associations with the classroom, how can they get involved? How can they really help us bring, these technologies or this innovation into the classroom.

Yeah, I love this conversation, and I think you're hearing a couple of consistent themes about learning in the future. One is the importance of agency and personalization. When we think about, and you heard this from all of our speakers, the kinds of experiences that are going to help young people find their purpose, help them navigate this increasingly complex world. They need to feel a sense of ownership over the world. their education, and they need to feel excited about learning. We released new research last

week with Gallup, where we asked students from all types of schools, give your school a grade, and the average grade was a B-minus. That's pretty bad. So students are telling us that they want an experience that reflects the kind of personalization, the kind of ownership, this sense that they are not just sort of in a transaction, but really building the skill sets and mindsets, as Allison said, that are going to help them be successful. The other thing that's coming up is that, you know, the pandemic really kind of changed the conversation

about where learning happens. And what we experienced is that learning can happen anywhere. Obviously, it needs to happen in school because that's where students are for most of the day. And also, many parents are thinking differently about what learning looks like, learning on a continuum, different experiences that provide students with those opportunities to engage and explore. And I think Minecraft is a great example, whether it's being used in a classroom or outside of a classroom, to build on top of the skill building and the mindset building that is happening. and we know that educators and families play a huge role in that and that the point of adoption for these technologies is too late.

So one of the things we need to be doing is incorporating early on in the R&D process, and I know this matters to everyone on the stage, educator and family perspectives when it comes to the development of these products. And I know the speakers will have great examples of how you're facilitating that kind of inclusive R&D, But as we think about really deploying these solutions across districts, across schools, that feedback loop and strengthening that feedback loop is going to be more important than ever before. Absolutely. And Arupah, maybe you can speak to that because maybe the group has already picked up on this,

but there's a level of hardware as well with VR, right? And so you not only have to get into these classrooms, but you have to get educators, school boards to opt in to this new cost. And so how are you doing that? How are you working with these folks? to not only open their aperture for the technology, but also the cost and figuring out how you can roll this out really broadly so that if it is as game-changing as we think it can be,

that everyone has access. Yeah, and I wanted to kind of piggyback on the role of the teacher before I was a math director for many years, and at NCTM, we've been talking about student-based learning, problem-brun learning, discourse for decades. So what's been really fun is that the role of the teacher isn't changing at all, she's just now able to do it. What was happening before

is we were telling them to do discourse, we were telling them to PBL, and the tools that she had were not amenable to that. So what's so great about going back to the question of adoption is when we go in, we're like nothing is changing.

You was the great teacher, you are launching an activation conversation, you put your headset on and they start working. Teacher has a web-based dashboard. So she's monitoring and providing feedback. Head sets then come down the last 12 minutes of class. She is architecting a beautiful conversation

that goes from the concrete to the abstract. So when we share this pedagogy, school boards are like, that's exactly how we've wanted to teach. How do we do that? Oh, by the way, there's a hardware investment. Well, how much? Well, VR isn't for every day. VR has a very specific role in the curriculum. All the other resources that we have, the wonderful computer software that's been built and evolved over the past couple decades. That's great. It just needs to be founded in something that's really real world and visceral for kids. So VR's only the first two days of a unit of study. Tells you why you're learning what you're learning it helps you derive a mathematical construct through a series of physical and tactile

experiences and then once that's done those card of headsets can go to the next classroom so we're deploying at about a 15 to one ratio at present as we create more and more content that might need to come down to 10 and potentially five over time but that's really been the cell and so I don't know if that kind of gets to the hard question but we're selling a vision that is not new people have always wanted to teach this way we are now providing the hardware and the hardware wasn't available five years ago. So for those you, I had a DK1. I'm a big VR enthusiast. And so VR tech, if you guys

have been following, a few years ago, it was not standalone. There was a bulky and vocal cord these big pieces of hardware. Movement tracking was bad. Hand tracking wasn't there. I track. It was uncomfortable at best. Right? That's not the case anymore. There's a psychological safety and a comfort that kids have. Most of them have

their quest at home. So when they come in and see they see the quest and the Picos and school, I cannot tell you if I had a nickel for the number of times kids that said, finally, you guys are coming into the 21st century because they get to see the tech that they have at home in their classrooms. Definitely. And I want to just touch on something that you've told me directly and you kind of touched on there. But like when you are going to these schools, maybe a misconception is that oh, teachers don't want this. But they're responding like, I've been wanting this forever,

right? And they're also something that I thought was interesting that you've mentioned is they are smart. And there's this maybe misconception around like, oh, can teachers handle VR? Can teachers handle AI? Yeah. It's like, absolutely. What happened with, like, many teachers who had held lower expectations for their kids of like, oh, well, like, we don't know if the kids can do it. Same thing with district leadership. The number of times the superintendent has told me, like, my teachers can't do this. I'm like, put your teachers in front of me.

And I'm going to say one example and then Taibo, but humble ISD, small $45,000 student district in Texas. They got, they were like, you know, only a small group of my teachers are going to want to use this. I said, okay, you know, because I've seen this a million times happen now. So I got 90 teachers in a room. They'd only purchased for about 40 teachers every single. single teacher put their hand up and said, we want to do it, and they were texting their friends. So 170 teachers opted into a program they'd only paid for for 40 teachers. And again, it's underestimating what our teachers are capable of and what they want. And as I

mentioned, like, kids are digital native. So when the VR heads come out, they're off to the races. They know the UI. So the teachers can focus on pedigoo. She isn't to worry about teaching them the controllers. The kids have that. She gets to focus on the questioning, the conversation, the convergence of the key mathematical notation conventions and outcomes. And we must build with that in mind that if we make teachers the conduit for the tech, we're setting ourselves up for failure. Something you also spoke to was the why now. So Saul, I'd love to get your take on this. I mean, maybe it's a little obvious with AI because the last year has been so crazy. But at the same time, tell me a little bit more about the why now. And specifically, what might this

unlock? Let's think a little more broadly about the next couple years and think past just like, okay, our kids will be using, you know, a chatbot. Yeah, I mean, to Romie's opening remarks, the situation is not good. And a lot of people blame the pandemic. But it was pretty bad before the pandemic. You know, people have looked at the NAEP results, and I think it was Detroit had 3% of kids, and I think there were eighth graders who were proficient in math, 3%. And it's like, wow, the pandemic was pretty bad.

Well, before the pandemic was 6%. So it was still pretty bad. And this is why you see things like a majority of kids in the U.S. who graduate from high school. So you're already talking about, you know, the top, I don't know, 80%. And then the subset of them that want to go to college, then now you're talking about the top 40 or 50%. a majority of them when they get to their colleges,

they don't even place into college algebra. This is the first time that mastery learning is being enforced and their colleges say, wait, you're not even ready to learn high school math yet, you're going to take remedial math and that has all sorts of predictive, you know, that the kids aren't going to graduate, they're going to have debt, no degree, et cetera, et cetera. And there's even been people say,

well, maybe we shouldn't make algebra requirement. I'm like, it's really like eighth grade level algebra that is considered a weeder class in our colleges. And that's because the average American student learns about 0.6 to 0.7 grade levels per year, which, yeah, after 12 years, you're at a seventh grade level. So that's the baseline. And so this is something that obviously we've, pre-AI, have been very focused on. And we have about 50 plus efficacy studies. I think we're the most studied platform because we're just out there. But they all say the same thing. Students put in

30 to 60 minutes a week. So not like a huge amount. They're accelerating 40, 50 percent. In some of our more, let's call it lab environments, we kind of run some schools. They're more like sister organizations where the students are putting closer to 30 minutes a day, because they're learning at their zone of proximal development, all of these ideals that ed schools have always been teaching, differentiated instruction, mastery learning, zone of proximal development, but teachers had no way of implementing it. We actually haven't been sharing this because we almost think that the results are outlanders, kids are getting about two grade levels a year, like when you do it consistently. And so it's not just genius kids who are doing calculus in middle school. It can actually

be maybe all kids. And that's at the same time that there's kids down the street who by the time they're 19, they can't do algebra, much less calculus. And it has nothing to do with, I think, any kind of innate ability. So that's the baseline. Generative AI, you know, there's always been this dream of, you know, I always cite these science fiction books, Diamond Age, where they favorite book ever. Oh, that's your feel very good. You're kind of working on it too, where we're working, taking a different, where it's, you know, a future where it's a neo-victorian China, where there's an app for educating a member of nobility's granddaughter. It's an AI-based virtual reality app that can educate a child and it gets pirated. It gets in the hands of these

200,000 orphan girls living in barges and they just take over because they get such a good education. And a lot of folks in the education world have always used the name of that app in the book, which is the young ladies illustrated primer. That's kind of like the code word for the true north that I think a lot of us. I mean, everything that all of us are doing are kind of somewhat inspired by things like that. And when, you know, the generative AI capabilities, not only did it throw a bomb into how our kids are going to cheat, et cetera, et cetera. So that created an urgency to figure out solutions. But it also, I think, laid out part of the solution,

which we can now get much closer to the Young Ladies Illustrated Primary. Yeah. By the way, for the educators in the audience, I would love if you could challenge your kids to come up with some positive sci-fi. We need more of that. Technology is, you know, it's inevitable and it's natural for us to picture the worst. I want to say one thing to that is one of my favorite books because the author is making a case for one of the most important qualities that young girls should have.

And if you were to take a guess, what would that be? If you're going to teach your young girls anything, what would that be? Yeah, the book makes the case for subversion. And it's incredible. It changed my life. I read it when I was a teenager before I went to college. So thank you for referencing that. You're right.

That's exactly what we're all kind of building towards. One last real quick thing is we kind of focused on some of the bread and butter stuff, but there's things you can do with the generative Aana that is you can talk to simulations of historic characters, fictional characters. And you can only imagine it's coming very soon, where it's not just chatting, you're talking. And then I think you start to merge these things with platforms like you're hearing, and it becomes completely immersive. You can go to ancient Rome. You could talk to Caesar. You could talk to Cleopatra. It's like the holodeck on Star Trek. And I honestly didn't think that was going to happen in my lifetime. I think that's

going to happen in the next five years. 100%. And it's exciting, or I think it is. In the interest of time, we only have a few minutes left. So I'd love to just give you all the same prompt. And we'll start with you, Romeo, and come back in this direction. I guess just any call to action, again, we have so many people in this audience, we have some students, we have some educators, we have a lot of parents, a lot of people that care about the classroom and perhaps how technology can play a role. And so just any call to action for people if they want to help out, if they want to learn more, Romeo, I'll start with you. This North Star and really optimistic and bold vision that you hear everyone on the stage talking about, it's not a political one. We want a generation of

kids to be upwardly mobile. And these are kids and families who have been trapped in generations of intergenerational poverty. And also, you hear it is not a clear path to classroom 2050. So we have to be able to hold the optimism and hope and boldness and also realize that there's going to be a lot of fear in the system pushing against this. So I would just come back to what I said at the beginning, which is the importance of collaboration in tackling this kind of educational change is just, we need it more than ever before. We need to be working across lines of difference. We need to be coming together around a common vision for what is possible for young people. And that's just going to take a different kind of conversation and dialogue.

And as we do that, as we talk about ed tech, continuing to come back to the cultural and structural changes in education, which means just, again, continuing to think about how we organize schools, think about the choices we're giving to families and making the job of teaching easier and more sustainable for teachers. Thank you. Allison? Okay, so from a gaming perspective, I talk to a lot of parents who are like, yeah, my kids love Minecraft.

I don't know what they do in there, but I know it's really, really lots of creativity. So my call to action would be to play the game with your kids. or if you can't play the game with your kids, sit beside them while they play because in Minecraft, they're like figuring out who they are as a human. Like if they're the person playing with their friends who's organizing the mission,

maybe they're going to grow up to be a leader or a manager. If they're the person who's designing like this incredible castle, maybe they'll grow up to have a world in design or urban planning. Or maybe they're the person that builds all the crazy machines with Redstone, So maybe they're going to be an electrical engineer or they're going to, you know, go into science in some way. And so it's a great way for you to get to know your kids on sort of a deeper level and then help them see the things that maybe they don't see about themselves. Amazing. We've got 30 seconds or so each. So Saul. Play with all of the above. That's the best way to learn it.

Yeah. And don't let fear get in the way of trying something out because if we do that, all that's going to happen is all of our kids are going to benefit from all of this. I'm not worried about them, but then the other kids are not. And then that's going to create a bigger divide. Thank you. Anna Rupa. I know this panel is called the classroom of 2050, but I actually want to be very clear. It's a classroom of 2023. It's happening right now. Go to a school near you.

All of us have deployed to hundreds of school systems. I mean, people who have been here for way longer than I have, millions and millions of kids are using. So I just wanted to like end with that, that the technology is here, the pedagogies here, the teacher training is here. We just now need to put our heads down and execute. And if I would have a last word,

the thing that I've seen missing in school districts, I've worked across many, many of them, I had very large teams. We were missing execution. And everything is now there. Pieces are there. We just now need to collaborate and execute together. Thank you. And thank you so much, everyone, for listening. Will you guys be around? They'll be around if you have questions. I'm so sorry we didn't get to audience questions, but thank you for joining us today. Thanks for listening to the A16Z podcast. If you like this episode, don't forget to subscribe, leave a review, or tell a friend. We also recently launched on YouTube at YouTube.com slash A16C

underscore video, where you'll find exclusive video content. We'll see you next time.