BEN:
And I had to explain, “No, you don’t understand. An education without data, which means an education without technology, doesn’t work. It is inferior to have a learning environment offline versus online if you design the digital learning environment correctly.”
REID:
Hi, I’m Reid Hoffman.
ARIA:
And I’m Aria Finger.
REID:
We want to know what happens if, in the future, everything breaks humanity’s way.
ARIA:
We’re speaking with visionaries in every field, from climate science to criminal justice, and from entertainment to education.
REID:
These conversations also feature another kind of guest: GPT-4, OpenAI’s latest and most powerful language model to date. Each episode will have a companion story, which we’ve generated with GPT-4 to spark discussion. You can find these stories down in the show notes.
ARIA:
In each episode, we seek out the brightest version of the future and learn what it’ll take to get there.
REID:
This is Possible.
REID:
You know, part of the awesome thing about doing this Possible podcast together is that when we address these really central topics, like education, we bring multiple perspectives to it. Each of us brings multiple perspectives. Say a little bit about when we’re reinventing education, what are the questions that you are thinking about that is important for us to address?
ARIA:
I’m like obsessed with education. It’s one of the causes that I’m most passionate about. When we’re thinking about education, it’s, “how can we serve everyone?” Because we’re not gonna lift up our society if we can’t serve every child, every student, every person. And so, a lot of people talk about the United States having the best colleges and universities in the world, but I would argue that the average student doesn’t get a great higher education in the United States. And I would argue further that I think we know that on the K through 12 level, we’re just not teaching young people for the careers of the future. I was actually really excited – so, as you know, I went to Washington University in St. Louis, and the chancellor sent out a note and said, “Hey, ChatGPT is here. Everyone’s talking about people cheating and how terrible it is, but it is the future, and so we need to embrace it and here’s how we can use it.” And I think on the flip side, what no one talks about is how can teachers use it? And how can professors use it? And how, like, we haven’t changed the teaching pedagogy in 50 years, and we have all of these teachers who have just been through the pandemic and… so exhausting. I can’t imagine what they’ve been through. We need to give them more tools so that they can teach the children of our future, the students of our future. And so I think the combination of technology and then empathy and compassion, both for students and teachers, is awesome. It’s a really cool combination.
REID:
Yeah. And it’s a 100% of, it’s part of what we’re doing here, in Possible, is like not to look at the technology as something that is just trying to outmode teachers, but to enable them.
ARIA:
Yes.
REID:
Because there are tons of these dedicated, sweat-blood-and-tears teachers who are trying to help kids get into their future. How do we help them? And that’s why we’re talking to Ben Nelson, right? The founder and CEO of the Minerva Project, the founder and chancellor of Minerva University. Ben combines a classic technology, Silicon Valley, entrepreneurial mindset with the dedication of “how do we take a first principle approach, revise what we’re doing in education – yes, use technology, but not just only technology.” Also, what’s the point of education? How do we bring people into it? Who, which kind of students you try to attract, how do you teach them? What does the in-person experience look like? What do the goals and outcomes look like? And to rethink all of that from a first principle for a modern society.
So as opposed to taking kind of just the canon of “everyone should just have a liberal arts education, and that is just the human right and dignity,” which of course he values, right? But to say, “no, no, let’s, as opposed to thinking, well, is the same education that was valuable in the 1940s the one that should be being done today?” Let’s think about how we change that at all levels. And that combination of Silicon Valley creative disruption, reinvention, technology, together with valuing this essential human institution. The kind of thing that we hope for every human being in the world, which is education as elevation.
Ben, welcome to Possible. We thought we’d kick off—because he is a close friend of all of ours—with the investment from Reed Hastings. And Reed and I occasionally get mail for each other.. [laugh] So each of us is like, “oh, no, no, you met the other Reid!” [laugh] But talk a little bit about that, that was awesome.
BEN:
Yeah, no, it was very, very exciting. I mean, Reed has supported Minerva University students for – this is his third year of supporting Mineral University students. And every year he has stepped up his commitment and it’s just a remarkable level of support. I’m always stunned when people, with no quid pro quo, nothing back, there’s no selfishness in it, it is merely an acknowledgement that he wants to see more students benefit from this ability to have this new kind of education applied and experienced all over the world. And I, you know, in many ways, it leaves me speechless. Right? But it is necessary, because if you want to create an institution that really isn’t there to just continue society as it is, but is all about social mobility and is about taking students that have, you know, they come from very challenging socioeconomic backgrounds and giving them an opportunity, right?
So to give you some perspective, a typical incoming class in the Ivy League will have about 13% of the students coming from a household making less than $50,000 a year. At Minerva, It’s 60. And the outcomes that we have for those students exceed the outcomes of Ivy League graduates, right? So, you know, but somebody needs to support that. Somebodies, right? So that’s just been, it’s just been amazing to see people come together and Reed being, of course, leading among them.
REID:
What’s the elevator description of Minerva?
BEN:
It depends how tall the building is. [laugh] But the very quick version of Minerva is: Minerva is built as a university should be, focused strictly on human development and learning. And so there’s no artifice, right? There’s no sports teams, there’s no campus. We select students from all over the world that share a common growth mindset. They’re talented, but they’re humble. They want to become better, right? We bring them from all over the world to live in San Francisco. We train them in 80 habits of mind and foundational concepts that form the underpinning of their entire education and, really, life skills. And then we have them live as adults in a residence hall in San Francisco for a year after which they will travel and live in six other countries over the next three years before coming back to San Francisco to graduate. And we’ve proven that this approach yields the best educational outcomes in the world, not just in theory, but in practice when applied to work entrepreneurship, even graduate school placement. Which is shocking [laugh] but true.
ARIA:
And so, can I ask, like, how do you do that? Like, I’ve studied higher education a lot, and I think a lot of people know now that it’s not the Ivy League that is the mobility, you know, it’s the CUNYs in New York, or the UT system. But 60% of your students being below $50,000 a year, like, that’s huge. Are you just cherry-picking the 20 best, or is there an ocean of other students that could be learning at Minerva?
BEN:
Oh, look, there’s a vast ocean. I mean, the reality is that, you know, if you were to go to any high school graduating class, not just in the United States, anywhere in the world, and ask them, okay, who here has heard of Minerva? I mean, in many schools you’ll have no hands raised, and then some you’ll have a couple who may have heard of it, et cetera. We’re still a tiny, tiny, tiny dot in the ocean of capacity. Every student that applies to Minerva, the exact same standard is applied to them. And it doesn’t matter what background they’re from, doesn’t matter where they come from, what country, what background, what ethnicity, right? What household income. We just, we just assess their merit. And that process yields diversity for free.
But then the real question is, what do we do with them? Right? Because that really is the core problem of education, and it really is at the heart of what a Minerva education is. And I think the easiest way to describe it is, if you were to think of a traditional education, it really is linear in its perspective, right? You have an A- or a B+, right? You have a 92 or an 87. The advanced ones: “You have competence,” “You don’t have competence.” But you know, we all know from day to day, this is completely meaningless, right? I mean, it doesn’t mean anything, right? What does it mean to be an A- skill level or B+? The real concept is: acquiring some piece of knowledge isn’t the end of education, it’s the beginning. Real education is, now that you know something, what do you do with it? How broadly can you apply it, right?
So, I use this example. If you have two students, let’s say you’re assessing them on a particular competency, and one student has a 92 average in a competency assessment, another one has an 87 average. But if the 92 average is derived from applying that competency in three contexts, and the 87 average is derived from applying in 48 different contexts, clearly the 87 average is better. And the education system not only doesn’t have a way to measure that, it doesn’t even acknowledge that as the goal of education. And when you create an educational system as Minerva has, that creates that as the only focus of the education, that’s why you get better outcomes. Because you actually train people to be wise, to be able to apply what they’ve learned in whatever situation they encounter.
REID:
So talk a little bit about that kind of first principles reinvention, with thinking about the kind of context of “how do I apply that education in my life, my work,” et cetera. What were some of the first principles that went into it early? What have you learned along that path, that you’ve gone, “here are key things that we’ve been learning that we hope spread widely?”
BEN:
So there are really two elements to it. The first is just how people absorb information. And education—and even in our own personal biases, we believe we learn when we hear or when we’re exposed. We do not. [laugh] Right? [laugh] So, and you know, it’s like-
REID:
Except if you’re listening to a podcast. [laugh]
BEN:
[laugh] Of course, of course. Yes.
ARIA:
Retain everything!
BEN:
Yeah, we’ll add techniques so people listening can actually figure out how they absorb this information.
But the real trick is you have to process. You have to think deeply. You have to think it through. You have to actually connect what you hear to other things that you know. Right? And if you do those two things broadly—think it through, and then make and use associations of particular pieces of knowledge to others—now all of a sudden you can retain what it is that you’ve heard or absorbed or come across. And by and large, universities don’t do that, right? By and large, you know, you sit in a lecture and you listen, and then a test comes and you cram for it, which is why you always forget what you passed on the test, just, you know, a week or two later. So one is a pedagogical approach, which, by the way, there are 40 years worth of research on this. We just compiled it and applied it.
But the second part, which is much different, I think, much more important and profound, is this concept of transfer. It’s the ability to learn something in context one, and apply it in context two, which does not come naturally to most humans. The example I like to give on this is, imagine you don’t feel well, you go to a doctor. The doctor would never prescribe medication and then say, okay, now let’s figure out what’s wrong with you. Right? Obviously, right. The doctor would diagnose you, you should figure out, okay, what’s the matter? What’s the root cause? And should have said, “okay, now let’s, like prescribe medication.” Right? But that same doctor, or the same patient, that understands everything about, you know, “oh yeah, I have to treat with medicine—you don’t want to just treat the symptom, you want to treat the disease,” et cetera. You know, they could go home after a long day, and they see the garbage isn’t taken out, and they get into a big argument with their spouse over the garbage. And how many people would say, “You know what, let’s just stop for a second… Why are we fighting over garbage? What’s at the root cause of the tension in our relationship? Let’s address that.” Same exact concept, right? But the principle has never been taught in a cross contextual way.
And that’s really the core of a Minerva education. You highlight the tool, the cognitive tool, make sure the student absorbs it, and then you allow the student to apply it over and over again, and you draw those connections such that they start to think that way as they go through life. So they think about when somebody makes a claim, how do I validate that claim? Is it a claim or is it a fact? When they propose a solution to something, they say, “well, is this a logical solution? Is it rational?” When, you know, they say, “oh, let’s solve a problem in some way,” say, “well, what is the unintended consequence?” Right? Or second or third order effects? I mean, these are all things that you can train the mind to do, but if you don’t train the mind to do it, it’s just not going to happen.
ARIA:
One of the things I thought was interesting as I read about Minerva, and this is directly related to that, is employers are looking for people who have more social skills, interpersonal skills, can work on teams, can solve problems. Colleges and universities don’t offer that in any way. And it’s something that you all focus on. And it actually reminds me of Opportunity at Work, which, Byron Auguste is a good friend of all of us, and he’s talking about how we don’t necessarily need college degrees, we need skills. Like, what would you say about that? The interplay between skills, college degrees, and the fact that universities aren’t preparing people for, you know, the future that we have now?
BEN:
You have to really reform the concept of university education, and you do have to connect what you learn in the classroom into real world environments. Work is an important element. Culture is an important element. Background. One of the great benefits we have at Minerva is that we bring students from all over the world. It’s this wildly diverse student body. And then they live all over the world. They live in seven different countries during their four years. And so they’re constantly recontextualizing what they learn, not just from course to course, but as they deal with their roommate, versus they deal with a person across the hall, with totally different approaches to dealing with issues or thinking through life. And then as they think about, well, how do I solve problems in a context of a startup city, San Francisco, versus in Seoul, versus in
Hyderabad.
REID:
The different life circumstances and the different communities obviously also helps teach the kind of multi-context problem-solving, the things that are important for entrepreneurship. Where is the drumbeat of technology in this? Just because it’s also an important part of how you changed—like what are the things where you’ve gone, “This has really worked, this was a little bit of a chimera I thought would work, didn’t work.” What were some of the technology elements?
BEN:
Yeah, absolutely. I mean, so the reality is, when you dig under the hood, I mean, it’s really easy to say, okay, we’re just gonna measure all of these things across context and all the rest. When you think about it in a traditional university setting, it’s science fiction, right? I mean, how can you actually track a student across 30 courses—that they will pick at random, right, I mean, they have multiple options, different majors, different electives, et cetera—and assess how they apply 80 different learning objectives across every one of those courses along four years, and understand exactly what context they’re asked so that you know when you provide an assessment, you provide feedback, how it’s connected to an overall learning map. Obviously, you have to do that in technology. And, you know, when we started Minerva in 2014 and said, “Yeah, yeah, yeah, we’re going through the great expense of bringing all of our students to live in the same building, and we’re gonna teach them all online.”
ARIA:
It sounds crazy.
BEN:
Yeah, it sounds totally crazy, right? [laugh] It’s like, “What, what are you…?” I mean, it just didn’t compute for people. Yeah. And I had to explain, “No, you don’t understand. An education without data – which means an education without technology – doesn’t work. It is inferior to have a learning environment offline versus online if you design the digital learning environment correctly.” And it was bananas. It didn’t make sense to anybody until COVID hit. Because as soon as COVID hit, all of a sudden people realized, “Wait a second, the education I’ve had offline; it’s just been delivered via Zoom. It is the same education. It’s just that when I’m on campus, I don’t bother showing up to class, and when I’m home, and I have to show up to class, I realize how useless it is.” [laugh] Right? And so the problem isn’t the technology, the problem is the nature of the product itself, the educational product. And what you can do in an appropriately constructed digital learning environment is first and foremost create mechanisms that ensure that people are engaged in their learning. Right?
So, as an example, part of Forum, which is the technology we built, is that we track the number of seconds that every student speaks in every class session. And we can color code their video, so that only the professor sees, to say, “Hey, you know, this student hasn’t been involved yet in class,” or, “This student is speaking too much.” And what they do is they then ride the average and they make sure everybody is involved. And so, that way, that’s something you just cannot do without the aid of technology. I’ll give you another, this is perhaps my favorite example: in our very first pilot that we did back in 2014, one of our professors, one of our founding deans, was Diane Halpern, who was, you know, a very eminent psychologist, former president of the American Psychological Association. She decided to run a small experiment. So she asked every one of the professors teaching the students to write down at the end of every class session who was the best student in class. And then she went back, and they did, and they just assessed using rubrics on what the students said. Everything they said, they just gave rubrics scores. You wanna know what was common about every single student that was marked as, at the end of class, this was the best student?
REID:
Talked the most?
BEN:
Talked the most, and male. [laugh] Every single student that both male and female professors said was the best student in class was a male. Every single student with the highest scores was a female. Every single one.
REID:
Wow.
ARIA:
So this is a way we can remove bias.
BEN:
That’s right.
REID:
Let’s start with something that’s already been part of the media discussion. And we’ll get to some of the GPT-4 stuff. But with ChatGPT, which is roughly 3.5, most of it is, “Oh my God, college essays, college applications going down in flames, stop it, ban it, stop the future.” There’s a little bit of, which I’m anticipating where you guys are, it’s like, “Oh my God, how do we use this? How do we amplify? How do we—we’re going to this future, what do we do?” So what is your guys’ reaction and your personal reaction been to ChatGPT, and what would you kind of offer to people to say, “Look at what you could do! Look at how you can change things!”
BEN:
Absolutely. So first, why did it take so long? [laugh] Because, because seriously, when I was starting working on Minerva in 2010, I was assuming, you know, that by the time we launched, this would all be there. Right? Because we kept saying, “Oh, yeah, it would all all happen.” We designed the program with that in mind. So as an example, we don’t offer any 101 or 102 level courses. They just don’t exist. We don’t issue credit for them, we don’t acknowledge them as university level courses, because the assumption is you can use some engine to get you whatever information you need, and you need to have a basic understanding of things in order to apply it effectively later on. And to me, that is really the superpower of ChatGPT, which is that—and you know, it was kind of an interesting paradigm that, I forget who told me this, he said, “look, when you do a Google search, it sends you on a research project. When you use ChatGPT, it gives you an answer.” And that’s just a fundamental paradigm shift. You, you need to be able to understand the difference between a fact and a claim. ChatGPT gives you a claim, right? And it gives you a very well-phrased claim, let’s put it that way. [laugh]
REID:
Yes.
ARIA:
It sounds authoritative. [laugh]
BEN:
It sounds authoritative. Yes. It sounds authoritative, it feels authoritative, and it uses a dataset of what is in the air, right? Yes. And so it gives you a popular claim. And what is more important than being able to train our students to both learn the authoritative way of presenting their claims, but also to be able to take apart those claims and see not the true or false stuff. The true or false stuff is easy. It’s the nuance. Right? And that’s the Minerva assessment bottle. It isn’t about black and white, it’s black and gray. There are certainly wrong answers in the world for any question, and you have to be able to discern what they are. But the real art is to see, in the spectrum, a potentially correct answer based on context, based on situation. Where are you going to go to? And my hope is that, you know, at some point, you see this, of course, in the science fiction movies, that ChatGPT will say, “Oh, there’s a 27% probability that, you know, taking this path…” Right? But that is what the world is. The world is full of, “That path is 27% probability of success, this is 24%, that’s 19%.” That doesn’t tell you what path to take, not even close.
And that’s the art of a real education. It’s to assume that you’re going to have that data and you focus effort and mind around that nuance. It also means that all of a sudden, all of the rote work that our students have to spend on compiling the information in starting off with answers will start – they’ll save all that time, and now they can devote that time, and now I can make the curriculum even harder, right? And so I can now graduate them even better, because they’re not spending all of that time that they did otherwise. I mean, I think about when I was in college and I was writing essays, what would I do? I would just highlight all of the quotes I liked in papers. I would copy and paste the quotes, and I would write my paper around them, and I would say, “How do I connect this quote to that quote?” And that was my ChatGPT version, right? But if I could just, you know, ask the ChatGPT to do that and then figure out, “Oh, well, you know, the quote you got wasn’t as salient as what I need, you know, find me something different.” Learn to hone that tool. That is what we’re gonna do in real life. So, embracing it and and then elevating standards, and this is really the core problem: how can you now make it better?
REID:
Exactly. Because like, “Oh, that’s a good argument about Plato, but here would be a better argument.”
BEN:
That’s right. That’s right. And how do you increase rigor? And in higher ed, we’ve had 50 years of decreasing rigor, not just in higher ed, in high schools. The education model has gone the exact opposite direction. Again, this is a problem of treating symptom versus root cause, right? People think, “oh, we’ve got an educational problem, we have varying outcomes, outcome levels, all the rest.” What’s the solution? Lower standards. You lower standards, then it’s easy for everybody to go to college, everybody to graduate. As opposed to saying, “No, no, no, no, no.” The issue is poor pedagogy, poor inclusion, you know, all of the things that don’t work. Rather than doing the hard work of addressing them, increasing demands on students, and giving them the tools to be successful, we say, “No, no, no, let’s just make it easier.” And that’s my big worry: that ChatGPT, you know, the devil’s bargain between student and professor, which is the less both of them work, the happier they both are, ChatGPT is going to enable that in a big way. So there’s one thing: to try and resist, which is futile and silly. The other way is to say, “Great, now I can do even less.”
ARIA:
So it sounds like one of the things that’s so magical about Minerva is the studying in seven different cities. I mean, I studied abroad, everyone like, has the best time. But you know, it’s both fun and it’s amazing for learning. Obviously not every student around the world can study in different cities. So we used GPT-4 and asked GPT-4 about what are the different ways that the future of education could be impacted and improved through technology like AI. And one of the stories was about a young kid named Malik who is able to use AR/VR to visit these different places to sort of get the Minerva experience without having to travel…
AI STORY:
Malik decided to use his VR contact lenses to do a virtual field trip to the Amazon rainforest. He had learned about the rainforest in his geography class, but he wanted to see it for himself and learn more about its animals. He logged into his school’s VR platform and entered the rainforest simulation. He was greeted by the avatar of his AI teaching assistant who spoke to him in Arabic, Malik’s first language, and what he spoke at home with his parents. The AI assistant told him that he had joined a group of his classmates who were also doing their book reports on the rainforest, and that they had a special guide from an indigenous tribe who lived there.
ARIA:
How do we recreate that amazing Minerva experience? Can technology help us? Can AI help us? Like, how do we do that for everyone who can’t travel?
BEN:
Yeah. So, I’ll say two things about that. I think first and foremost, we have to be both sober about the opportunity and the costs of real live experiences compared to virtual ones where obviously, you know, there is a trade off. But we also have to understand, well, what is the delta? I’ve given I don’t know how many millions of “VR in education” presentations and, you know, one of my favorites, like, “Oh, look, on the computer you can see the pyramids and you can like go through them!” And I was like, “Well… I don’t think it’s like being at the pyramids!” [laugh] And maybe you get there at some point, but I think there is a real tangible value in the unexpected. In the struggle of, “Huh, I thought I would be able to go in a cafe and use wifi because I have an important meeting or a class at this time, and I showed up and it doesn’t work.” Right? “There’s a cow that’s crossing the street, now I cannot.” I mean, there’s basic things that you can experience that are important. So, to me, I would actually change the framing on that. Which is, “What is it that you get by this travel experience? And if you cannot change your context of city, how do you wind up changing your context where you are?”
So I think that you can have seven different—radically different—experiences in San Francisco. Think about what you could do, and in fact, even in any metropolitan area, right? What does it look like to actually be immersed in a very economically stricken, poverty stricken neighborhood versus what does it look like to be in an immigrant enclave versus what does it like to be in the middle of the financial district? And then to actually live in that context. And all of a sudden your cost is a bus fare. There’s something about that experience that can be replicated. You can also think about it not just from a cultural immersion. You can think about, well, how can I experience different workplaces? And here is an enormous advantage of technology and of the COVID awakening, right? I will say that before the pandemic, it wasn’t easy for us to convince employers to take Minerva students as interns while they were studying halfway around the world. Now it’s a given. And the ability for you to be in one place and work for multiple different companies, have different contexts, and work on projects. The barrier of place has evaporated. There’s so many ways in which we can leverage technology and leverage opportunity of context if we think about: what is the ultimate goal? Because the ultimate goal is transfer. It is context change. And anything we can do to create authentic context change, I think is crucial.
REID:
How did that GPT-4 story stimulate your thinking on the possibilities for what we can shape in the future?
BEN:
Yeah, absolutely. I mean, to me, here’s what’s remarkable the most about, you know, the whole GPT and AI thing. It has passed the Turing test. Like, you cannot tell that someone didn’t write that story. And it has clearly been trained by what is out there. [laugh] So, even the scenarios of what happens to these students, what their assignments are, they’re pre-GPT assignments. [laugh] And so it’s like, how do I use this incredible tool to do the same thing I did in a world before this tool existed? [laugh] And I think, to me, all of these stories that were generated highlight how crucial it is to envision a different baseline. And again, to start from the ground up, to start from first principles and understand, look, now that you have enormous increased productivity, the ability to really process things, how do we redefine what it is that you expect from students, even at young ages? How do you expect them to actually demonstrate that they’re using these tools not to achieve the old goals in faster ways, but to actually create new goals? And to me, that was the overarching theme of all of this. And I think that that’s really the big opportunity for educational institutions: it’s to up their game. That will scare a lot of institutions. But I think that’s, really, to me, the big takeaway.
REID:
Completely agree with the upping the game and it’s go to the opportunity, go to what can be so much better, versus fearing the change. I mean, that’s why we have this word “luddism.” It’s like, “No, no, actually a steam engine and a loom is a good thing.” Like, “Yes, people can still handweave, but we all get much better clothes across the entire world if we have a steam engine together with a loom.” Same thing in education. Now, one of the comments that you made when we were chatting a little bit before, I think is very important to share with folks, which is: the other reflection, when we were asking GPT-4 a bunch of questions about how to improve education, you were like, “Well, those are the classic higher ed education points, because it’s trained on the current stuff that’s in the internet.” Right? Say a little bit about that, as it were, blind spot in the lens of the training, because that was part of the judgment about how to use these tools well and how to not be misled by them.
BEN:
I mean, one of these stories, which was one of my favorites, talks about how a girl that comes from a far, rural area of Canada winds up getting her masters in linguistics. What is a better talking point for a university and for nobody else? [laugh] I couldn’t imagine. I mean, you know, it is one of these absurd lenses where, rather than talking about, “Here are her capabilities, here is what she was able to do, how she was able to contribute to society,” the end point was, “Boy, she got that master’s degree!” And I think that that’s really the shift. We’re working on – one of the things that Minerva does is that we help establish new universities around the world. We help establish new university programs, even within existing universities. One university, actually, one I’m going to later today, I’m heading down to Mexico, is being established in Mexico. And as we were talking to the folks who were starting it, they were very blunt about, look, they do want to create entrepreneurs and people who will start new businesses and they said, you know, “Look, we want our students to come and go through the entire process, but if they get 80% of the benefit halfway through… okay!” And they’re ready to go, go! And because it’s the end, it’s the goal, right? And if somebody can do it faster and they can achieve that, you don’t need the stamp. And for an institution to have that kind of thinking, that’s the thing which is going to move us forward, that is the thing that is going to get us to understand. And, by the way, when we understand that our collective responsibility is to have an enfranchised citizenry that understands the difference between a fact and a claim, that knows how to have correct expectations of what can be done as opposed to be fed fantastical concepts and ideas, we are going to be a much better society. But we don’t think about it in those terms. We generally say, “Oh, you know, we need people to think this way versus that way.” You can’t win that argument because not everybody thinks in a particular way. But no one would say, “No, no, no, no. Nobody should have, people shouldn’t have those tools.”
REID:
So one of the things, of course, about the whole AI as amplification intelligence or augmentation intelligence is of course, it writes out a first story about what, what an outcome could be. You look at this and you say, “Actually, in fact, this outcome story could be so much better.” What would you do? Adding to it, changing it, modifying it, looking at it with the judgment to say, “No, if it changes, it changes, it changes—this is the society we are trying to move towards. This is what’s possible.”
BEN:
Well, first I’d take two steps back. I’d ask ChatGPT to figure out what’s her purpose, right? Not what’s her goal. What is her purpose? What’s her motivation? Why is she enamored with language? Why is it that she wants to be able to communicate in different ways? Why is she attracted by the world beyond the circumstances in which she found herself born? And then I would try to navigate—and shows my relative amateur way of programming ChatGPT—but I would try to navigate it towards going back to how that purpose is fulfilled, less so about the milestones and more about the capacities that she would need in order to fulfill that purpose. And the more that you think and guide, whether it be the AI or even the learner, towards that framing: Why is it that you wanna do what it is? Now what is the core problem you’re trying to solve in order to get there? What’s the gap between what the current state is and what the various solutions are that get there? And then you go through it systematically. Now you get into the gray zone, now you get into the world where, “Huh, I now have a plausible path to maybe solve a problem.” And now let’s compare that to other plausible paths and see what’s best.
ARIA:
So at the end of every podcast, we’re going to ask a few rapid fire questions. So could you tell us a movie, a song, or book, some piece of culture that fills you with optimism?
BEN:
This is gonna be a very strange response, but it is The Life of Brian by Monty Python. [laugh]
REID:
“We’re all individuals!”
BEN:
“We’re all individuals!” That’s right. And what I love about The Life of Brian is here’s a movie—it was made 50 years ago—and every scene in that movie was relevant then, it’s relevant today, and it was relevant 2000 years ago. [laugh] And why does that feel—most people look at that and despair, right? We’re dealing with the same thing! It fills me with optimism because, overall, humanity does progress. And we progress with all of our insanity, and all of our flaws as humans, and we do get better. And technology is, whether we like to admit it or not, the fuel that enables us to get better despite all of our issues. So The Life of Brian, you know, it’s: “always look on the bright side of life.”
REID:
So on a different vector, where do you see progress or momentum outside of your industry that gives you optimism about what’s possible?
BEN:
I’m a big, big believer, strangely enough, in augmented reality. I think we are very far away from cracking it, but, right now, our augmented reality is the iPhone. You know, we always think about how much the world progresses. And I like the fact that—let’s assume that progress, we just measure this century, so far, you know, 2000 to 2023. And now let’s think about the most recent benchmark, 1900 to 1923. Think about those 23 year spans. What happened to humanity between 1900 to 1923? My grandmother was born 1900, my grandfather was born in 1890. I remember my grandmother telling me the first time she saw a carriage that was not drawn by a horse. Wow. And she didn’t know what was going on.
ARIA:
Wow.
BEN:
Flight. World War I. The pandemic, the collapse of monarchies, the introductions of representative governments, which in 1900 was almost unheard of, right? In 1923 it was commonplace. I mean, we were a little bit away from discovering penicillin. Just the rapid change in that 23 years. Now, fast forward to—just go back. Imagine somebody in 1923 going back to 1900, or vice versa, how radically different of a world it would be.
What is different about today versus 2000? The difference is the internet comes with us. [laugh] The internet existed in 2000, I was there. [laugh] It did a lot of the things that we do now, like we had food delivery to the home, that existed. [laugh] But now we have an augmented reality, and, frankly, AI is finally becoming real. Starting to, but we haven’t seen the applications yet. And so, to me, the big bet is that AR is going to be the vehicle that is going to enable all of this change. I just don’t know how. [laugh] I don’t know if it’ll be a glass display, I don’t know if it’ll be a contact lens, I don’t know if it’ll be auditory, I don’t know if it’ll be some other thing, an implant. But I fundamentally believe that if we can effectively augment our reality with AI-inspired information, the superpower will be with us everywhere. And that is an opportunity that—I don’t think anybody has started to fathom what it could mean for humanity.
REID:
Other than a few science fiction authors.
BEN:
Perhaps. [laugh]
REID:
So, can you leave us with a final thought on if everything were to break in the right direction for education in human’s futures over the next 15 years, what would that future look like? And what are the steps that we need to be taking in order to get there?
BEN:
I think it comes down to the what, the who and the how. The “what” is that education will embrace a completely redefined centering around the ability to teach people how to apply broadly what they learn versus demonstrate recall. Without that shift, nothing else happens. The “who” will, again, if everything breaks the right way, all of a sudden we’ll reverse a 50-year decline in education mobility in, at least, talking from an American perspective, and reverse it dramatically and fast. And just to give you some perspective, in 50 years, the top quintile of the socioeconomic distribution doubled its college completion rate or college attainment rate. The bottom quintile increased by only 50%, and it increased from an extraordinarily low base. And so you would see a dramatic shift in who gets educated in this way.
The “how” I unfortunately think requires public policy. And, to me, the easiest way of doing it is very simple, and which, by the way, I think is a policy that people on the right and the left would gravitate towards, which is the government should declare that any university can educate whomever they want. But any university that doesn’t have, at least, a proportional representation of the socioeconomic distribution of the country among its undergraduate student body loses all non-profit status, loses all tax benefits, has no access to research dollars, has no access to NIH, or any other government funding of any kind. And I guarantee you, every research university in the country the next year only enrolls students from the bottom 50% of the socioeconomic distribution. They will never take that—maybe they’ll have a few people right from the top to, you know, pay for the buildings. That’s it. And it’ll be a dramatic shift in social mobility in this country.
REID:
Wow.
ARIA:
I love that. [laugh]
REID:
Yes! [laugh]
ARIA:
Ben, thank you so much for being here.
BEN:
My pleasure.
ARIA:
So wonderful to hear about Minerva and everything else that’s going on. Really appreciate it.
REID:
Always amazing. Thanks, Ben.
BEN:
Great to be here.
REID:
That was awesome. [laugh]
ARIA:
So amazing. What I found so interesting—and I have known Ben for like 10 years, and you read all the articles about Minerva, that’s about, you know, we’re out-leading Harvard and we’re the most elite college in the world, but there’s no one who came off as more concerned about equity and lifting everyone up and, “How can we radically change society so that we can increase social mobility?” And I thought that was awesome because there’s just such a rhetoric of—and I even fell into it, like, “Oh, these students, they won’t be ready for Minerva or they’re not…” And he was like, “No, you just need to raise expectations and all of us will benefit.” The kids at the top will benefit, the kids at the bottom will benefit. And so just that idea of raising expectations across the board was just a real takeaway for me.
REID:
Yeah, and by the way, that ties very much to the kind of area that I was thinking of, which is I, all the way back to a decade plus ago I wrote about how technology is going to reinvent the diploma and that it isn’t just going to be a degree, but a set of skills and competencies. And I’d already reflected on a lot of the places where our educational system is a mismatch for life, because work is mostly a team sport. How do you work together? The educational system which overly emphasizes your individual achievement, sitting in, taking an exam by yourself versus the “How are you collaborating and working together?” And when he was working through what were the key things like, “No, no, you can get that from Wikipedia, you can get that from GPT-4, you can get that from ChatGPT.” Then come into the social work circumstance, work on its context application. How you do it together? That’s the thing that you’re gonna amplify. And so the education thing isn’t the, “Are you really, really good at reproducing a compare-and-contrast essay?” It’s a “Is that context of life and work going to be there?”
Like, I hadn’t realized how much the Minerva—sure, you start in San Francisco, you end in San Francisco, but you go to these other places and the other places are part of that context shift. Part of that, “We’re treating you like adults. You are being amplified into adulthood for this.” Not kind of the delayed “We’re all partying at university.” But as ways of doing that. And that obviously ties directly to that: no, no, no, using these kinds of patterns, we can take the very broad brush of talented folks who haven’t been trained on the, “Oh yes, I’ve been learning to do a compare-and-contrast essay since I was nine with my private tutor” but like everyone in order to learn those, those right outcomes and that the outcome is navigating a good life.
ARIA:
I mean, obviously a lot of our discussion was about equity and ensuring that everyone has access to it. And I know you’ve thought a lot about the digital divide, and so how do you think the digital divide with AI will affect the future?
REID:
Well, this is one of the questions I’m going to want to come back to you on, too, because this is what I always kind of think about. In the future of education, one of the things that Ben pointed out, which is a fundamental truth in all societies, which is that the structure of whatever the elite class is—generally has more money—kind of operates to try to capture opportunity for the children that restricts social mobility. But that question about how do we enable as much of the great talent as we can to express themselves and contribute to our society is part of what I love about the aspirations, the American Dream, always be building towards it. Part of why I ended up in entrepreneurship is as a way of doing this. How did you reflect on it? What was your, like, “Ah! Here is something to really grab or double down on, or imagine the possibility of?”
ARIA:
Well, I loved both Ben’s radical proposal of making the elite universities reflect society and their student body. I also thought it was really interesting, so, we had three AI stories that we looked at and we actually, we didn’t talk about it, but we had the story of Wes Moore, the current governor of Maryland, who has an extraordinary story, you know, grew up in poverty in the Bronx. And he was lucky enough that he had this mom and these grandparents who sort of pulled him, scraped together all their money so he could go to private school and then military school. In the three AI stories, it was all about young people who got super lucky and got a scholarship and super lucky and got this…. And what I liked about Ben’s answer is “we need a systems change.” This idea that only if you get lucky, or if you happen to be low income, only if you are plucked out and get this amazing—it’s like, no, that has to be wrong and this system has to change.
And so when I think about AI, it’s like how can we use AI to stop people from being lucky? [laugh] How do we give the same opportunities to everyone? Part of it is personalized education, part of it’s the changing of the schools, part of it’s the teachers being better. All of those things. We just need to stop this being “lucky.”
REID:
Well, I don’t think you can stop luck. I think luck is a thing. [laugh] But the question would be: how do you not make it that the only exceptions from the less economically advantaged classes and people is only luck? How do you get as much of that, or how do you get as much broader base for, as opposed to narrow luck? As opposed to, like, you roll a six-sided die and it’s only when you get a six that it works. How do we move that to three to six? Right?
ARIA:
How do you expand luck?
REID:
Yes. Because there’s always some fortune. Like, you’re born in the US versus you’re born in a war country or something else. There’s always some of it. But I think part of human progress is how do you broaden it? It’s one of the reasons I like you know, Bob Wright’s Nonzero book so much. How do we make more of our systems positive sum, where we’re creating a better and better future? Part of the reason why we’re doing Possible. [laugh] Right? And that’s a very important constituency, because, by the way, it’s many more people. So if you can actually, in fact, make it so that—because talent across a million people is distributed across a million people. So if you go, “Well, we only got a million here and we got a hundred million here.” Well, how do we take the hundred million, and how do we enable absolutely much more of that talent to the benefit of everyone? To themselves, too, but the benefit of everyone.
ARIA:
Absolutely. No, couldn’t agree more.
REID:
Possible is produced by Wonder Media Network, hosted by me, Reid Hoffman, and Aria Finger. Our showrunner is Shaun Young. Possible is produced by Edie Allard and Sara Schleede. Jenny Kaplan is our executive producer and editor. Special thanks to Surya Yalamanchili, Saida Sapeiva, Ian Alas, Greg Beato, Ben Relles, and the team at LinkedIn Media Productions, including Justin Blumenthal, Hasan Ahmed, Keith Cheng, Eric Carlos, and Gabe Lomeli, Jr., for video production on this episode.