Jeremy Rossmann of Make School on Income Share Agreements and the Future of College

Okay, Jeremy Rossmann, welcome to the podcast!

Jeremy: Glad to be here!

Interviewer: You are one of the cofounders of Make School. So, for those who don't know, what is Make School?

Jeremy: Make School is a new college in San Francisco. We offer a bachelor's degree program in applied computer science. It's a college in the ways you'd expect a college to be—a college with dorms and a campus and full-time faculty. We're located near Union Square. What's really special about it is that students don't have to pay tuition unless they get a job. The curriculum is entirely project-based, and our faculty come from industry backgrounds. We have a lot of partnerships with the companies that our students want to work at, and they can get a bachelor's degree in an accelerated time frame. So, you can actually graduate with a four-year bachelor's in two years if you take our intensive track.

Interviewer: And how does that intensive track differ from the normal four-year program? Because when I went to NYU, I was pretty much working all the time on school stuff. So, how do you guys get it done in two years?

Jeremy: When you look at how colleges typically structure their academic calendar and even a given week, what happens is a lot of students are basically going to lectures here, and they are going to labs here and there, with a lot of dead time in between that's self-directed. We want to be preparing our students for a transition into the workplace, and we've realized that the best way to do that is actually to structure their week like a work week. So, students are typically expected to come in at 9:30 in the morning, have what looks a lot like a morning stand-up with their individual coach, who will be their coach for the rest of the program, and then spend the day working on their projects, attending certain classes and labs. When you add up the time that students spend all day, every day in that forty-hour school work week, plus the time they spend outside on homework—which for us is all projects, there's no traditional problem sets—you can award students enough credits that if they take the course load that we recommend and they come in and they're spending 40 hours a week in the school and working outside on their projects, they can get enough credits to earn a bachelor's in two years.

Interviewer: And how are your students trying to differentiate themselves from people coming out of traditional CS programs?

Jeremy: We are all project-based, and we equip our students with portfolios that basically speak for themselves. From the very first weeks at Make School, even in your CS theory class—which almost sounds like an oxymoron—but our CS classes are project-based and applied; you're going to be building and launching small projects and then larger products. By the time you make your way to being in front of a recruiter, they can see a tremendous amount of evidence of your coding ability, the skills and technologies you're familiar with, your theoretical background, your practical background, because that all is visible in the portfolio that you've built.

Interviewer: We recorded an episode with Austin from Lambda School and, in listening to a couple of podcasts with you, doing research for this episode, you started talking about parents. It's something I had never really thought about with Austin. I think perhaps the people who are attracted to Lambda School versus Make School are later in their career, maybe taking about switching. What is it like for you dealing with the parents of eighteen-year-olds at this point?

Jeremy: I've come to love it. I'm usually the point person on the team for parents. It is important to note that our student audience is majority students who would otherwise be in college or who are college age. So when people, you know, we just had a prospective student weekend last weekend—100 students coming in from all over the country, parents coming in from all over the country—and the questions are, you know, it’s Make School versus Georgia Tech, Make School versus Stanford, Make School versus, you know, a community college or a state school, not Make School versus a boot camp. Because our student audience is really looking for, you know, they're going to a college; it's a question of which college, or they're looking to transfer to a better college.

So the parent questions often have to do with why is this different, why is this worth it, but also how can this possibly exist where you don't pay tuition unless you get a job? The parents have more of that. You know, they’ve been exposed to the professional world. They're trying to do the math in their heads—like, how can there be a campus and dorms and how can all this exist if they're not charging upfront? Right? And you know, there are two simple answers to that question. One is that we’ve been running the income share-based program since 2014, so we actually have five years of data. Our students are getting jobs at Facebook, Google, and Tesla and all these great companies, so the salaries are there, the payback is there, and it's really working well. The other is that we, you know, we get financing to cover our cost of educating students. So we're the school that actually goes in debt when students enroll—it's kind of a reversal. Normally, you know, you go into debt when you enroll in college. At our college, we go into debt, and we hold the risk, and you only pay us later if you get a job. But we're on the hook for that debt anyway, and that puts us in a position, and that's what parents really appreciate when they realize it—that we have no other incentive than to be providing the kind of education that's gonna get you a successful career outcome.

Interviewer: Let’s finish the program!

Jeremy: At the end of the day, do you find that that's the core driver for parents? Or is everyone just going in because, like, job is the most important thing? Because, you know, that historical thing is like “Oh, expand your mind, teach you how to think”, etcetera.

Those things are incredibly important, and actually, we rediscover a lot of those same liberal arts concepts and values when talking to employers. So if you say, okay, Make School's primary goal is to get students access to top software engineering jobs that are typically only hiring these days out of top universities, that’s the reality, right? Everyone tries to get their funnel out at the MITs and Stanford’s of the world. So, if you want to go to employers that hire there and get them to consider Make School students, what you discover is that even they are going to tell you that a lot of those liberal arts educational experiences are part of what makes a candidate stand out.

So even if you were just narrowly career-focused and you rediscover that actually a lot of the well-rounded education, the critical thinking skills, communication skills, and cultural skills that come with a liberal arts education are things that employers do value. But we have to realize that the pure liberal arts education that is explicitly not career-focused—that doesn't make any claims about whether it can be relevant to a job later—that's an education that was pioneered for and by elites, and any student who's considering attending a school that offers that kind of education needs to be conscious of the fact that the flagship schools, the Harvards and the Yales and the Princetons, they do practically don't have low-income students. They do a lot of PR about the scholarships they offer and all this, but the reality is anywhere from sixty to seventy-five percent of the students who go to schools like Stanford and MIT come from the top 20% of families by income. Oftentimes, you get up to a quarter of the student body that's from the 1% of families by income.

So they don't get career training, they learn the liberal arts, but that's really like a luxury education where you don't have to worry about your career because you're from the 1%. Of course, those give me opportunities lined up. If you went to Harvard and learned about culture and literature, we serve a student body that is demographically much more diverse. We get students who transfer in from top schools, we get students who come from wealthier families, but by and large, the majority of our student body comes from low to mid-income families. It's a demographic almost inverse from what you see at Stanford or MIT, right? And you simply cannot, with good conscience, provide a luxury expand-your-mind-only education to a student who is not in a position to have any of the privileges or sort of freebies that you get from being a one-percenter who goes to Harvard.

So it's not an either/or; it's really an and. But we start with a focus of how do we introduce a technical education that is both theoretical and practical and all project-based so that students are employable? And then how do we ensure that five to ten years down the road, in their careers, they're just as competitive as a student coming out of Harvard, or Stanford, or MIT? We do also have a liberal arts component to education that is integrated with the technical education.

This is related to one of the questions that was sent in about being an autodidact. So, William Trischka asked, "What are some ways to encourage greater independence and autodidactic behavior in the students pursuing technical skills or knowledge?"

So in the context of, okay, we're gonna teach you whatever, like parse and jQuery and whatever you might need now, but how do you ensure that you're teaching someone how to teach themselves technical skills?

So this is another thing that's very important to employers, right? They don't want people who have just learned a trade, who know how to assemble Ruby code today, and then when the framework has updated next year, they're kind of lost. Or when the company makes a shift to a different framework, they don't know how to relearn. So we do a number of things to ensure that students are really strong in that front.

The first is we do have a robust theoretical computer science fundamentals curriculum that students have to go through to really understand kind of behind the hood, like under the hood, what is going on, and how do all these concepts work together? How do these different paradigms and structures underpin the different technologies and tools they’re learning how to use? We also teach the concepts in sort of repeated different contexts.

So if you only learn programming in one language or one framework, it's hard to actually differentiate in your mind what is a property of the language, what is a property of the framework, what is a property of best practices, and what's actually a theoretical CS concept that underlies it all. When you start seeing this across different technologies, different languages, different frameworks, and you've had that theoretical CS education behind the scenes, then you're able to pick out and understand what is core to a specific framework, which is what is a general and universal concept.

As a simple example, we see a lot of our students taking our iOS development courses and then getting jobs as Android developers, right? We see students learning a number of stacks for web development and then getting a web development job in a totally different stack in language. We're seeing that again repeated across data science and machine learning. There's a number of different packages and tools that different companies use, and onboarding onto a new one is not a foreign experience by the time a student has completed their Make School education.

Interviewer: And I’m kind of curious about this word in general. So, what percentage of the world do you think is an autodidact anyway? Like that might be a false premise.

Jeremy: I think it very much is. While it is absolutely important to be a school that is augmenting students' ability to be lifelong learners and making sure that they're future-proof, the way that you do that, I think for most students, has to be through mentorship, dedicated curriculum structure. What I found actually was that I was not an autodidact, and I failed over and over and over again to teach myself computer science. I actually tried for the first time in eighth grade. My dad gave me a book. He's a huge autodidact; he taught himself everything out of books. "Jeremy, here's a book on Visual Basic." And I just couldn't do it. But I was actually a very high-performing programming student because I was lucky to have really amazing CS teachers in high school—actually, two really amazing CS teachers.

What I learned then was that probably the majority of people who could be awesome software engineers are not autodidacts. And if you look at courses that are kind of self-served like the MOOCs, you see that it's industry standard to have two, three, four, five percent completion rates, meaning that probably 95 percent of people are actually not a fit for the type of education that is autodidact first. So what we need to balance is recognizing that actually, when we look at industry and we see that the best programmers we know taught themselves, but we're probably seeing a survivorship bias. We're seeing the fact that there aren't really good structured forms of education that enable people who are not—who are like me, who weren't good at teaching themselves—to actually be excellent software engineers.

So at the other end, you see an over-representation of self-taught autodidact software engineers that almost has biased the industry toward that profile, when in reality, when you look at the data behind a lot of different educational programs and you look at, frankly, what our students are doing who landed these awesome jobs, I'm now certain that the majority of people who could be a great software engineer at a really competitive and selective company—the majority of those people are not by default autodidacts who could just go to a MOOC and get all the information they need. So that's why students come to us, right? An 18-year-old who's looking for their first higher educational experience generally wants to go to a college because they recognize that that structure is going to accelerate them, and it's going to be better than just trying to do something self-served online. I think that addresses the majority of the population.

Interviewer: Yeah, I feel the same way. Like basically, everything I've taught myself has been just enough to get the project done. But then if you were to throw me into like a Google production-level environment, it would be a disaster! When I tried to deploy our website for the first time, because in the early days, we were going through Y Combinator in 2012. I was also writing curriculum; when basically, one of two instructors, my co-founder and I were the two faculty. I was also a web developer. We were gonna deploy a simple Django website. Even at that level, even though I'm an MIT dropout, I’m like decently clever, I could not get this going without a mentor.

I hired a Django developer to tutor me for 30 minutes a day during that week that we were going to get the first website off the ground. My speed of learning and ability to actually be high-performing as a software engineer was literally 10x if I had a smart human available in real time to notice where I was blocked and unblock me as I got stuck and ran into conceptual hurdles.

That's one of the things that, you know, I’m most glad that we can provide to our students. Just seeing that happen live! A lot of folks who frankly have accumulated a huge amount of self-doubt because with that survivorship bias, you look around in your CS class, right, and you see that, you know, maybe the faculty is not such a good teacher, the curriculum is old-school and traditional, it's not working for you, it doesn't match your learning style. But there’s that one kid who's still crushing it, and you go, "Oh, must not be for me! Right? I can’t be like that!" Right? They’ve been coding since they were 12; they just read online, and they're fine!

And actually, if with the right teacher, with the right curriculum, you know, project-based learning is in the context of computer science the best way to activate people's passion. Because it’s like the reason you're learning this is because you have an idea that you want to see exist in the world, and you're learning so that you can launch the thing, right? That's what every single class is about! That kind of structure can get you to the level of the person who's been learning online since they were 12.

But people often don't realize that because they look at the industry where it is today, and they kind of work backwards and draw the wrong conclusions. It's just like the fact that you're not sitting next to someone—you don't realize that even the best programmers get stuck, like setting up an environment for 10 hours all the time!

All these little things that you bash your head against the wall, and then you sleep on it, and then you realize it.

And you asked me, like, how can you do a bachelor's degree in two years? Another thing is just from a pure competency and speed of learning perspective. With our faculty in the room who have been engineers in Silicon Valley, they’re there, and it's all project-based learning. So they're not there to lecture to you. They're not there to just be a transmitter of information—the information is already available. We have our curriculum; you know, it's all flipped classroom-ish. A lot of it is available to students, I mean, in the form of written materials that we produce.

So the faculty's job is actually to be the kind of the live tutor, the live mentor, the unblocker, and that is, again, for most students, a 10x acceleration in the speed of learning because you shouldn't be spending 10 hours setting up your IDE, you know, when you're starting a new framework, especially as a student! What a waste of time, right? You should be spending 30 minutes and then learning all the concepts, and the conceptual stuff that you need to know to be successful.

Interviewer: Yeah, no, I mean, I got stuck learning Ruby and then I often attribute Python to my learning. But I think actually it was learning Python while I was friends with people who were much more technical than I was, really helping me debug.

So a question, like going back to the parents really briefly, you’re in an interesting position where you kind of have like two stakeholders making purchasing decisions, right? How do you go about communicating to both the potential students and the parents?

Jeremy: For the potential students, the message is very much, you know, from the heart: this is the college that I wish existed. I was a student who was at the cusp, you know, learning how to program at the cusp of the app store revolution. My co-founder and I, in our junior and senior years of high school, were amongst the first to start teaching ourselves how to make iOS apps and launched iOS apps. That experience was so fulfilling that when I got to MIT, I was pulling all-nighters to work on my app with friends and not on my problem sets. That gave me some insights into what are the motivational factors that drive a college student to focus on their education.

So what our message to students really resonates because it mirrors my experience and my co-founders' experience, I think really mirrors what a lot of students are experiencing, which is a stagnant, you know, style of education and the in-classroom experience being very old-school, with a really significant set of questions about whether what they're learning in class is gonna translate to what they actually want to do later.

For parents, it's different. There's some overlap. Parents latch onto the fact that, you know, I'm a college dropout who started a college, so that's an interesting thing for them to wrap their heads around. For students, it's like, "Oh, that's really cool! Right? He was at MIT, he thought of something better." My co-founder topped out at UCLA; we teamed up and we started this.

For a parent, they want to know—I think they're thinking a bit more long-term, so they're worried: is my child making a short-term decision? Is this a shortcut? Are they gonna get all the fundamentals and foundations that are gonna set them up for success later? Or should they fall back on a more traditional option because that’s what the parents are comfortable with?

But what I will say is that there's been a huge shift in the last two to three years. Parents are way more aware of the student debt crisis, the ROI challenges in higher ed, the fact that a huge number of college graduates are underemployed. So whereas two to three years ago, a lot of my conversations with parents were dealing with skeptics, I now get more conversations, proportionately, that are parents asking how they can advocate—who say, "We’ve been freaking out as our child is now a senior, or they’re at a college accruing debt, and we're realizing this is not the way it was when we were going through college. This is not a world where you can take a summer job and graduate without debt. This is not a world in which the economy is booming to such a degree that everyone can have a middle-class life out of college."

Right? That you need to actually be tactical about the higher education you choose to make sure you don’t end up in one of the many traps that that parents have seen their older children or other folks fall into over the past few years. So, you know, I had 30 parents visiting campus this weekend, and whereas before it was a lot of questions about legitimacy, now I think the biggest questions are about future-proofing, just making sure this isn’t a shortcut and that the integrity of education is still very high and not at an opportunity cost to something that might set a student up for longer-term success.

We have to discuss how the project-based learning that we provide actually encapsulates a whole bunch of future-proof skills because when you teach computer science in a project-based way, you've got to do some design, you've got to do some user acquisition, you've got to do some customer discovery, you've got to work on a team, and actually, you get more communication skills than you might in a liberal arts communications course, and so on and so forth. These are a lot of the questions that parents have that we have to address.

Interviewer: Yeah, it seems like a lot of these learnings, right, it’s just like one generation learns it the way that they were brought up and then they got their jobs, and then they teach it to their kids, and then their kids get screwed.

Jeremy: And for many, many years, that teaches their kids and then it worked, and over the past 20 years it stopped working.

Interviewer: So, what do you think, like for instance, like if I had kids? I don’t have kids, but say we had a kid today, and they were one year old. By the time they’re at college age, what do you think this environment looks like?

Jeremy: I think that the landscape has dramatically shifted by 18 years from now. There are a few forces at play. We are, I think, the first of what I hope to be many new, and what has been a market that has traditionally been dominated by incumbents that have not faced any new pressure from new schools. One of the things that’s really fascinating when I talked to parents and students is that it takes a few times of repeating for people to be able to visualize that there is a new college in San Francisco, but you have a campus building. We have these beautiful dorms; there's full-time faculty. People are not used to the idea that a new college could be born because colleges are something that just would have been there forever.

In any market where you have the only players that have been there forever, they've been stagnating and they have not been facing pressure to improve their efficiencies and their quality. So I think 18 years from now, there's gonna be an increase in pressure on traditional higher ed to be aligning their incentives with their students.

As more and more students tune into the fact that if you just copy the default path to success, which is try to get into the highest-ranked school that you can, and then trust that if you follow your passion that you will get an education that will set you up for success—that as more and more students realize that's just simply not true, there's gonna be a demand for alternatives. We are actually one of the first colleges to be created under new accreditation rules. We were accredited this past fall; we offer a real, accredited bachelor’s degree, accredited by the same accreditor that accredits Stanford and UC Berkeley and so on and so forth.

They created these new policies in 2014, and what that means is we’re relatively early in what is potentially a new wave of colleges being born all around the country. I also think that the financial structure of college is gonna have to change. Students also start tuning in to the fact that the incentives between them and a college like MIT is not right, and for students who really are passionate about research and academia, things are working a little bit better.

We have deep respect for students who want to go that direction and are very close to a lot of folks who are in that world. It's awesome to be dedicated to the furthering of fundamental knowledge and to be doing research and preparing for that. But, you know, a lot of surveys show the majority of students' primary concern is at the end of their bachelor's: are they going to be employable?

So, you know, in the face of that primary concern, that's going to drive a lot of decision-making, and that's gonna put a lot of pressure into colleges. We’ve started seeing, actually, that other colleges, even big ones, are adopting our income share agreements. The largest example is Purdue. The documents that we helped co-author back in 2014 when we piloted our first college program—which has turned into this accredited bachelor’s that we offer now—those same documents, as they've iterated, have been iterated on over the years and are now being used by a number of other colleges.

What we’re starting to see is that income share agreements are not just this thing that applies to programming schools and Make School, the new CS College; it's actually higher ed is taking notice and starting to roll it out at a larger and larger scale. I think, 18 years from now, you’re gonna have options that look like the options we have today, but it’s going to be a lot more obvious what they are.

Harvard is going to be more obviously recognized as a luxury good, as an expensive purchase that works for the elites, and it’s something that if you can afford to get your kids there and you can afford to not focus on hard skills, then great—you’ve purchased a luxury education. It's gonna be more obvious what is actually tied to real-world outcomes, and there’s gonna be schools that are increasing, gonna increase their level of innovation in curriculum and also in financing to prove to students that they're putting their money where their mouth is and that their incentive is to help students achieve their goals.

Interviewer: So do you envision they will also shorten their programs like you have, in doing so allow for people to basically make a stack? So, you're like, "Okay, I'm gonna buy fancy Harvard for two years to drink, and then I'm gonna go to Make School and actually maybe learn?"

Jeremy: I think that students—we’ve actually seen this line of thinking with students, like the idea that they're gonna stack it tends not to play out. Once they have an opportunity to exit education with a career or a next move, they’re generally not gonna look for education again for a little while. So instead of a stack, what I think it’s gonna be is this lifelong learning pattern where students are gonna enter the workforce a bit earlier, they’re gonna work for a few years, and then they’re gonna return, whether for a master’s program or a second bachelor’s that’s gonna refresh their skills or maybe expose them to something that has emerged that was not a big thing at the time that they graduated.

You know, if you graduated college five years ago, you might be wanting to go back to really learn what’s fresh and new in machine learning, for example, right? And so I think that kind of is gonna repeat itself, and we’re gonna see students, you know, entering the workforce earlier but then returning for further education at a higher rate. But I think that’s not gonna be stacked. It’s gonna be, you know, years down the line.

Interviewer: So what about degrees in general? So Vikram Malhotra asks, "When are we going to do away with degrees?" You know, for instance, like I saw this article—I didn't read it, but I saw the headline as we, exactly, like Google and Apple do away with college requirements or something to that effect.

Jeremy: Companies, you know, here and there do PR around this idea that they're dropping college requirements, and we’ve been on the ground talking to companies every week since 2014. We had our first inaugural sort of college program, and the reality on the ground is quite different. Top schools, the ones that attract a majority of wealthy students and have a lot of prestige associated with them, continue to be the major hiring funnels for new grads.

For these companies, including Google and Apple, a large number—at some companies it can be almost all—but a large number of new grad hires are typically sourced from internship classes, meaning students who, while they were in college, interned at that company and then came back for full-time. I can tell you with certainty that there is no university recruiting internship program that I've come across that has dropped degrees as a requirement in any sort of systematic way.

Maybe there's been a one-off exception here and there, but the reality is university recruiters have a territory of universities they recruit from, and they're only there to work with people who are in degree-granting programs. So, on the employer side, we haven’t seen this play out nearly commensurate with the amount of PR that's been associated with it.

Interestingly enough, though, several of our students—without degrees—who went through our program before it got accredited have landed at Google and Facebook and so on and so forth, so there is some truth that like it's not a hard requirement. It's a question of, what is the proportion of their workforce that actually is representative of that shift in policy?

Although that should be important to note, this is just for CS, right? That's what most of your data is also true. It's probably less true otherwise. So while it might be strictly true, as I've seen with my students—after being in the program, I have two alums who are at Google and either have degrees, or I have two alums here at Facebook, neither have degrees because they were in our program before it was degree-granting.

It is true, like you can get jobs now; you can get full-time jobs at these companies without a degree. But if you look at the proportion of the workforce that's actually being sourced in that way, it's not addressing the fundamental equity problem, which is there is no structural pathway for low-income Americans to access higher-earning software engineering jobs at large tech companies.

There's a substantial number of anecdotes of people who come from low-income backgrounds who are there, but nothing that even makes a small dent in the statistics. When you think about the fact that 100 companies are going IPO this year in Silicon Valley, we are having billions of dollars of wealth transferred to the founders, investors, and employees of companies who all, on average, come from the top 20% of families in the U.S.

What I want when we talk about 18 years from now, what happens, I want us—and hopefully other schools like us that are educating students who are lower to mid-income—to have it too. Such that when there’s the next wave of 100 IPOs in the next boom cycle 10 years from now, that that wealth creation accrues evenly to folks from all backgrounds and not just to the folks who are kind of the incumbents who've had access to elite education and who had access to elite job opportunities.

So, yeah, it's technically true, but if you look at it sort of on a societal level, it's not moving the needle.

Interviewer: What's the same thing for Y Combinator, right? Like everyone thinks, "Oh, well, you know, Bill Gates and all these people dropped out of Harvard, so therefore I need to go to Harvard and then I need to drop out, and then I'll create a multi-billion dollar company." The fact is I was an MIT dropout going through YC at the age of 20, and I was basically one of like five of my profile—way below average in my entire batch. People thought, "You know, I would meet people who thought, oh yeah, you are the prototypical YC founder: you are an MIT dropout, 20 years old going through YC." Now what I tell them, actually, there were only like five other people in our batch out of over a hundred people who were, you know, younger than like 25.

In fact, my co-founder and I were the only young founders in our batch whose company is still alive. So when you look at it from that perspective, you look today at what are the companies from my YC batch that still existed, are still successful. There are zero examples beyond Make School of a still-existing company that was founded by a college dropout who was young.

So, yeah, where I see people kind of have mistaken perceptions. I think the other part of the question is whenever we do away with degrees, is less of a practical question about, well, are companies dropping the requirements and so on and so forth, but more of like a higher-end question, right? Are degrees still relevant? Are we going to increasingly have to circumvent degrees because so many degree programs are dropping in relevance?

Yeah, and there's a bit of an arms race here. I think in a lot of the country, it is totally true that degrees are dropping in relevance and students are looking for alternative routes of education. I think there’s gonna be a pop in a sort of higher ed bubble where a lot of colleges are gonna go out of business, and it’s gonna be a lot of pain at higher ed because of this factor. So Vikram is on point about that.

But at the same time, there are gonna be folks like us who're gonna try to make degrees relevant. We're trying to essentially bust this false choice between should I get a degree or should I get relevant skills, right? It's like, why not both? The answer has often lied—well, accreditation standards, for some reason, hamper your ability to do both. The good news is, at least in the West Coast, our accreditor, WASC, is under new management. Jamie Studly, the new president at WASC, created these new regulations in 2014 that we've fallen under to become accredited and be able to offer accredited bachelor’s degrees.

It involves, you know, partnering with established institutions as a quality control measure; it's a really thoughtfully put together set of rules. The end result is now there isn’t that false choice, and we’re a great example of it. I think if we can pull that off, if we can be a fully project-based school—like some of our faculty don’t even have master’s degrees, right?

Things that people think are not possible as an accredited school, yeah. But it turns out that the accreditor in the West Coast is very open-minded to this and understands they see what we see, which is all the things that Vikram sees, and makes them ask that question: are degrees still relevant? And they want to be relevant.

The last thing I'll say on that point is our views shifted on this tremendously as we started talking to and better understanding who our inner student audience is, who came from low-income backgrounds. To say that you can succeed without a degree is on the face of it true, but it is most true if you come from privileged—not just privileged, but as a college dropout, I went to a high school that is in the richest ZIP code in the United States.

When I look at my buddies who are also college dropouts and doing okay, we have something in common, which is that we all come from relatively privileged backgrounds; we had a safety net. We have a safety net. So when you talk to students who are lower-income, especially parents of those students, you realize that to tell somebody who is African-American or Hispanic/Latino or otherwise underrepresented in tech who might be the first in their family to go to college or the first to get a credential, they don't default belong in the way that we’ve structured this industry, which is a big problem, and they don’t have the safety nets and the privileges and the networks like I got—a professional network out of my high school that is as good as what people go to get Stanford MBAs for, right?

That’s totally unfair! But if you're like 90% of Americans, that's not you. So this notion of degrees no longer being relevant ends up really applying to those with privilege—the 2-4 percent we discussed who are autodidacts and who can bypass all sorts of traditional education and still emerge skilled—and it leaves behind basically 90% of Americans.

Interviewer: And so, what about the schools now that exist that aren't necessarily technical? You know, Evan Ward asks, "Should liberal arts colleges consider adopting income share agreements?"

Jeremy: I think that you know, the advice that YC would give to any organization is talk to your users and make sure that the product you’re building is a fit for your users' needs. Universities don’t do this; they should do a lot more of it. But other organizations do pull college students, and it is known that the majority of college students, even at liberal arts colleges, primary concern is, am I gonna get a job after graduation?

Americans want to live the American dream; that’s very clear, right? And so, is there a mechanism to align incentives between the college and the student? Also, to be able to be more open about what is a program that actually is compatible? I don't think it’s Jesse. Yes, I think it’s actually widely applicable to a whole bunch of fields that have lagging student supply relative to industry demand. It could be everything from medical fields to teaching to software engineering.

Any program that you cannot apply is, uh, you have to ask yourself, is this program really a good fit for the 90% of Americans who don’t have that safety net who are looking for something that will better their lives? Or is this actually a program that's more of a luxury good that should have a pay-upfront price tag?

I think it'll make it a lot easier to know actually what you're getting into. It'll be very telling when a college says, hey, you can go to our nursing program, our education program, or CS program, but, oh, but, you know, English majors have to pay this much upfront. That tells you something right about what their confidence level in being able to deliver on the American Dream for the students in those schools.

So I think it’s gonna turn out to be widely adopted in spirit. I don’t know if it’s gonna look exactly like income share agreements everywhere, but the key feature of ISAs is you don’t pay if you’re not successful, and there are ways to replicate that with loans through loan forgiveness as ways to do that with income share agreements. The course structure is gonna matter.

Do we want to see the higher ed adopt this idea of aligning your incentives and not having students be paying you if the student is not successful? I think that there may even be more innovation in how that is structured exactly, and other schools may do it in a way that is not ISA. But what's important is that more and more higher education institutions should absolutely be looking at what outcomes they provide and be starting to make that their number one priority because it is their students' number one priority.

Interviewer: Have you seen instances of, like, predatory ISAs?

Jeremy: Yes, to some degree. There's definitely folks who are using it in a way that makes us nervous. We're like, "Okay, yeah, you’re not disclosing appropriately to students what their future payback burden might be, things like that." We also see schools use it as a way of actually reducing the amount of scholarships they give.

Here’s how that might work: you have a school where this tuition is $40,000, and on average students pay 20K upfront and 20K in scholarship. In most schools, by the way, there's no such thing anymore as students paying less tuition. Schools keep track of this thing called the discount rate, and basically, most schools don’t actually charge the list tuition price to almost any students, right? So, let’s say the average student is getting a twenty-thousand-dollar scholarship and twenty thousand dollars in tuition.

Some schools are realizing that what they can do is they can replace the twenty-thousand-dollar scholarship with that twenty thousand dollars as an enrollment fee. It feels similar; you only have to pay twenty K upfront. But now, what used to be scholarship money is actually extra revenue for the college. You haven’t expanded access really; you’ve increased the total cost of education for the student when the whole point is actually not to do that.

You are not aligning your incentives with your student because you’re still getting half of it upfront. So I think ISAs, or similar structures, are most powerful when schools do—it’s kind of like the bet-the-house model on it. Yeah? When they say, okay, we’re really not gonna take enough money from you upfront that it’s worth it for us. It’s only gonna be worth it for us if you are successful after graduation. Schools that do intermediate models, where they supplement with paid— I have mixed feelings about that, even though we do let students mix and match if they want to.

Ninety percent of our students go full ISA, and that's what we want. We’d love for that to be 95%, not 99%, because we want—we want to be, every day, waking up and thinking our only priority is not what is the fancy new renovation for the multi-billion dollar campus building, what is our sports team doing, how is our research going? None of that. We want to be thinking every single day: how do we help students be successful?

Interviewer: I think you’re awesome, and something that only like a startup would have tried all out.

What do you think about the students who come in who want to meet entrepreneurs?

Jeremy: We get a lot of those students. At a prospective student weekend, we had a hundred students in the room, and we had a guest speaker who was like, you know, "Show of hands if you want to start a company someday," and it was like three-quarters of the students in the room!

Well, we saw those students as, hey, I was a young college dropout entrepreneur. There are almost no people like me out there. All the successful entrepreneurs I know look like some in their late 20s, early 30s, or older. My father himself is on his like eighth startup, and it's folks who have had significant exposure to building and launching products. So when they understand a problem that they want to solve, they have experience in doing customer discovery, you know, crafting an MVP iteration cycle, getting feedback, making sure that they're building something that actually has product/market fit, and then it's folks who have had experience in the real world enough to expose them to categories of problems that are worth solving.

The challenge, if you're a college student who wants to start a startup just out of college—which, of course, there will continue to be students like that, and that’s awesome—but the general challenge is you haven’t had exposure to building things. So, your startup might be your first rodeo in building something, which is, you know, higher risk.

And the problems that college students face are limited to problems that are faced by a very narrow segment of the population—with no money! Most of the purchasing power is held by people who are between the ages of, like, 22 and dead.

And most of the problems in the world that really need solving are not necessarily visible to students in college, but we do, to mitigate that, is obviously a whole education is about building products that solve problems. The entire culture of the school—from every brainstorming phase to every welcoming taku orientation and so on and so forth—is about pushing students to see software as a creative tool to make an impact in the world and solve real problems.

We’re always asking students, even in their CS theory class, right, how can you leverage what you’ve learned in the projects you’re building to actually be solving a concrete problem? That forces students to go out and already go through essentially test cycles of being an entrepreneur where they have to conceptualize a product, they actually have to launch it; they actually have to acquire users for it.

Doing that in the context of college, I think it sets our students up for success as entrepreneurs later. And my suspicion is that two to three years after graduation, when students have had a couple of years of work experience, we’re gonna see a lot of our students leaving to start companies, and I'm excited for that because when they do, they'll probably come back to us to hire, and it’s gonna be great for the ecosystem. We’ve already seen there’s already, I think, three groups of students starting companies at various scales out of the program.

Interviewer: Awesome! So just in closing, we have a new batch of YC coming out. Yes, so you're an alum; your company is still alive. Yes, well, one or 2012, and now it’s—right, way longer! What would be your advice to someone going through the batch?

Jeremy: I think that the anti-pattern that I saw in my match and in many matches was to think that you are above the generic foundational YC advice. The generic foundational YC advice, as I remember it, is make something people want. Be darn sure that what you're making is something people want. Focus: don't be doing fake work. Don't be going to conferences unless it really impacts your bottom line. Don't be taking coffee meetings left and right. Don't play entrepreneur; execute to a high degree of fidelity based on what your customers want.

Once you get out of YC, another thing that we see is, you know, students playing optimization games around funding and that kind of thing. You know, way more startups that I know than I would have expected came within weeks of running out of money. The other advice that I think emerged a little bit later in my YC life, which was this notion—are you default dead? Are you default alive?

How aware of that are you on a weekly or monthly basis? Being very, very clear on how, at a strategic level, you're going to steer your company to ensure that you're shifting from being default dead to default alive is probably one of the most significant things you can do. So if you're making something people want, if you're focusing only on that and not the other stuff that seems like work but isn't, and you're shifting your company towards default aliveness instead of default deadness without trying to optimize too much on the terms of your fundraise and really trying to optimize for not dying, I think you’ll do great.

And the biggest tragedy is to really be onto something and then, through a tactical or strategic sequencing mistake, end up running out of money, basically, or having to do a fire sale when you could have changed the world.

Interviewer: Awesome. Thanks for coming in!

Jeremy: Thanks so much!