The Future of Education, Student Debt Crisis and Unbundling a Legacy Product
Also, Lambda School, Income Sharing Agreements, Automation, the next Mayor of NYC
Some time in the future, I would like to explore the idea that we can redesign education entirely with a mind toward maximizing actual knowledge absorption. The model of lecture-based education is not exactly based on any sort of cognitive model about strategies that maximize learning. Lectures are, basically, just designed to maximize the number of students who hear an idea, as opposed to optimizing for learning. In academic circles, this strategy is pejoratively referred to as “transmissionism” and there isn’t really anyone who thinks it’s an optimal strategy. It’s not that lectures don’t work - they are just far from an optimized model for maximal learning. I imagine that the future of education involves a minimization of audio-based acquisition of information combined with deeply active engagement with material/problem solving. As far as I can tell, the academic world is already beginning to move in this direction, but the hundreds of lectures I sat through during my freshman year pretty clearly show that we haven’t gone nearly far enough. I would love to hear/read about any work that is being done to build the future of knowledge transmission/acquisition based on actual empirical evidence of effectiveness - if you have resources/knowledge on this, please reach out and share your thoughts. This isn’t the idea I’m exploring today, but it is really interesting and, if I feel deeply informed on the topic at some point, I’ll write about it in the future.
Incredibly thoughtful technologist/researcher Andy Matuschak wrote some really interesting stuff about reimagining books and the lecture-based model of education that I suggest reading if you’re interested. I hadn’t thought before about reimagining the concept of a book from first principles but after reading Matuschak’s work I can’t stop thinking about it.
Apologies for the delay in publication - I was neck-deep in college applications and needed to make my final push the last 1-2 weeks before they were due. I have been germinating on this post for a while, ever since I read The 100-Year Life. The 100-Year Life is not directly about education, but it set me off on a mental journey, exploring the deeper ramifications of extensions in lifespans and accelerations in technological progress, and I ended up in the world of education.
Before I get started, and while we’re on the subjects of learning and the acceleration of technological progress, go read this interview with Stripe founder/CEO Patrick Collison. I consider myself an intellectual curiosity zealot and Patrick Collison has to be one of the most intellectually curious people in the world. I could seriously read/listen to the guy talk about absolutely anything. He is probably the most thoughtful, broadly+deeply informed young person in tech. Go check out his thoughts on tech, innovation and pretty much everything else. When people talk about wealth redistribution, my mind often goes to Patrick Collison. There’s no one in the world I trust more to allocate capital to societally-beneficial, ROI-efficient causes/companies than Patrick and, frankly, I’m excited for a time when his Stripe equity is eventually liquid and he can start putting some of that capital to work, pushing the world’s progress forward. Plus then I can get my hands on some of that Stripe equity people are always talking about. I am seriously deep down the rabbit hole of exploring the links Patrick provided throughout the interview. It’s pretty unbelievable not just the amount that he reads but the amount of really high quality information that he has readily available and is constantly sharing. A subscription to a list of “things Patrick Collison is reading” would be worth more to me than a subscription to any traditional media outlet could possibly be.
While we’re talking about the fuel behind my techno-optimism, check out this post from Sam Altman (ex-president, Y Combinator.) Moore’s Law was the driving force behind a ton of the technological progress of the 20th century and the early 21st century. Moore’s Law will likely continue to drive technological progress for the foreseeable future, unless you buy into the veracity of pessimistic warnings of the asymptotic future of compute power innovation.
On to the (irregularly) scheduled programming:
Unbundling Education
In 1995, on the IPO roadshow that launched the dot.com bubble, Netscape CEO Jim Barksdale famously proclaimed that “There’s only two ways I know of to make money: bundling and unbundling.” It’s become an oft-used Silicon Valley trope, but it hints at a very real cycle in many domains. A nice example that has unfolded in recent years occurred in television. Over the past few years, networks like HBO and massive media companies like Disney have pushed the unbundling process as the traditional cable subscription model falls apart. Led by Netflix, it feels like every major network/media company has launched a streaming service by now, and each of them is un-creatively named with a “+” at the end. I’d urge you to watch over the next few years with a consideration toward the cycles of bundling and unbundling. As these services proliferate and compete, winners emerge, and streaming quickly becomes a more expensive option than the simple cable subscriptions they replaced, one competitive strategy will be to bundle multiple streaming services and offer a discount for signing up for the bundle. That will especially be necessary among lower-quality streaming services, more than it’ll be necessary for Netflix or Disney, but it will happen throughout the market. Sounds an awful lot like a cable subscription, doesn’t it?
The student debt crisis in America has become a favorite political bogeyman in many areas of the political spectrum, but it’s also a very real, widespread issue holding back many young Americans from beginning to build their savings and their livelihoods. The discussion centers around loan sizes, societal pressure to attend college and the bankruptcy-proof nature of student loans, but the solutions don’t really go farther than “have the government pay it off” or “make college free.” I don’t want to debate the merits of either idea in this piece, but I do want to spark a conversation around the root cause of the student loan crisis.
Like all bundled products, a bundled education - the traditional college experience - makes sense for a lot of people and provides tremendous value to many students. I would not advocate for a significant reduction in opportunities to enroll in a traditional, completely-bundled university. However, like most bundled products, it is a tremendously inefficient offering for a student who does not find value in the majority of the products/features available in the college education. If you don’t care about interscholastic athletic competition, you can find a college without sports, and the cost structure may be a little bit cheaper for it. But if you aren’t interested in, say, student organizations, or dorm living, there are not tremendous options in place for you to find a university where you get everything except those features. Because there are tremendous barriers to entry in the field of (respected) higher education, and because of innovation-averse mindsets in many areas of academia, we see a dire lack of iteration on the traditional college experience. What I want to see is an explosion of offerings that unbundle the features of the traditional college experience so that students can choose significantly more affordable educational options without sacrificing on their priorities.
What might these new-look schools entail? One example would be Lambda School, a professional development program for aspiring software engineers. Lambda takes the technical education, uses cohorts to create some sense of community, and turns the career services up 1000x, and strips away all the other, traditionally bundled features of an education. Lambda is not considered a college replacement, and, as far as I can tell, their students don’t consider it one. But there is no reason that it couldn’t be. Because they have stripped away so many of the aspects of college the many students enjoy, but some may consider frivolous or unnecessary, they offer a product, the education/job placement, with a significantly cheaper cost structure than would be possible under the traditional educational model.
Maybe another set of students love collaborative projects and outdoor recreational activities, but don’t care for lectures, traditional career services, normal university facilities or any of the other features of the traditional model. Someone could innovate with a school that focuses on project-based learning and is based in a beautiful part of Colorado or Montana where students can explore the area around them while focusing on the parts of academics they find most valuable. Again, the cost structure in this university would be significantly lower than traditional colleges.
Aside from increasing optionality and the availability of varied, specifically-tailored choices for students, innovation in education would create the downward pricing pressure that could halt the wild inflation in college prices that has driven the student debt crisis. With this downward pricing pressure, universities around the country would be pushed to look inward and make a real value calculation of each of their offerings and programs, rather than continuing to bloat budgets with the understanding that students would pay for the increasing tuition. There is no easy path to get over the barrier of respected credentialing that protects legacy academic institutions, but if someone like Lambda School figures it out, it will push universities everywhere to improve in ways that are hardly imaginable today. This isn’t about hurting the traditional educational model, it’s about improving it and helping to create a sustainable long-term model for effective, affordable college experiences.
Automation, #YangGang and a Longer Working Life
As I mentioned, I recently read The 100-Year Life, a critically acclaimed book by two London Business School professors that made the shortlist for the 2016 McKinsey Business Book of the Year Award. The premise is that, based on the trajectory and historical consistency of improvement in the human lifespan, someone around my age has about a 50% chance of living to 100. I’m in no position to assess the accuracy of the actuarial calculation here, but I found the concept and the authors’ initial exploration of its implications to be incredibly interesting. When people regularly live to 100, it won’t make sense to retire anywhere near 65. When you combine longer life expectancies (and with that, longer implied career windows) with the constantly increasing pace of innovation and automation, it becomes very clear that the contemporary three-stage life structure is not going to work in the new world.
In the three-stage life, everyone is expected to go through an adolescent period of education until they’re in their late teens/early twenties, a longer “career” period for 40-45 years and a retirement period from about 65 to the end of their life. This isn’t the exact path that everyone follows - grad school, high school dropouts and early/late retirements can all change the picture a bit for some individuals, but the framework holds for the population at large. In the near future, a 15-20 year education period followed by a 40-50 year continuous work period is just not feasible for the majority of the population. Skills that are learned in the education period simply won’t be useful over the entire working period of most people’s lives. Not nearly enough vocational skills are learned during that educational period in the first place - more on that below.
The three stages feel like a fact of life when it’s the only societal structure you’ve ever known, but it’s actually a fairly recent development. Teenagers and retirees are both really creations of the 20th century - before then, life expectancy wasn’t long enough and education wasn’t good enough to justify anything other than childhood and adulthood, with a little education followed by work until the end of your life. It’s hard to imagine a change or an addition to the stages of life, but it’s not nearly as radical or as unprecedented as it may initially feel.
While I found the premise to be exceptionally thought-provoking, I was fairly disappointed by the choices the authors made in writing The 100-Year Life. I felt they were almost maniacally focused on the insane savings rates that would be needed to pay for a longer life with a longer retirement, and they didn’t spend enough time focusing on wider societal ramifications of longer lives and longer, more convoluted career paths. In a world where the existing formal education is noticeably insufficient in preparing young people for the careers ahead of them, our resilience in the face of technology change depends on the rise of alternative forms of education which we’ve only begun to explore. I’d like to present a model which I find to be at least one worthwhile option to “fight the battle” against automation that has been highly touted by a dark-horse presidential candidate who looks like the next mayor of New York City...
When Andrew Yang talks about automation, he leads with Universal Basic Income. UBI is the policy that propelled him to the main stage of American politics and it’s still the acronym with which most Americans identify him, to the extent that most Americans have heard of Andrew Yang in the first place. To hear Yang’s name, think “that’s the UBI guy” and move on is to miss both the nuances of his argument and the lesson that should be taken from his techno-prophecy. Yang is decidedly pro-business and pro-technological progress, as a former tech entrepreneur himself. I think he could run the government more effectively and more efficiently than anyone else in or around politics, and he certainly would run the country with better KPIs than anyone else in politics, but I’m not interested here in making the Yang Gang argument. When Yang speaks outside the debate format, and has a chance to explain the nuances behind the buzzwords he uses on stage, he speaks both excitedly and ominously about the coming disruption that will be brought on by innovation. In many parts of the country and sectors of the economy, the tidal wave of innovation has already wiped out countless jobs - there aren’t a lot of auto workers in Detroit these days. The example he always uses are truck drivers - it’s a well-paying job, tough on physical health but otherwise very steady, and self-driving technology is coming for the entire industry. A common, slightly offensive and all-too-dull refrain is to tell the truck drivers to “learn how to code.” It’s nice to imagine that the issue were that simple, but tens of millions of Americans whose jobs will be wiped out by automation are not going to be the software engineering workforce of 2035 (though I hope some of them will try.) The need for a mass respelling of America is imminent, and it’s far from a one-time issue. It will be a constant battle to maintain employability for each member of the American workforce and each person around the world as innovation continually knocks out traditionally secure jobs.
As a country and a human race we’ve faced incredible challenges before and technological innovation continues to be an overwhelming force for positive change. At each new innovation, there will always be people who cry out about the damage that will be done and the horrors of technology. Writing was once going to destroy each child’s ability to remember facts and stories. Socrates notoriously hated the practice of writing and refused to write anything down - were it not for Plato following him around and documenting his wisdom, we would never remember his name today. In more recent history, luddites destroyed textile machinery in 19th century England during the Industrial Revolution to protest the loss of their old jobs. I’m fairly sure they landed on their feet, and the world was better off for the Industrial Revolution. This too shall pass, but it’s about time we start preparing for the massive need to reskill our population over the coming decades.
The Current State of Economic Mobility
According to Stanford research a child growing up in America in the bottom quintile of income has between an 8% and 9% chance of reaching the top quintile in adulthood. If that child grows up in certain areas of the country, that chance drops to about 5%. If they grow up in Silicon Valley, they have a nation’s best 12-13% chance of reaching the top quintile at some point in their life. This difference between geographical areas is generally attributed to differences in quality of education. Having grown up in Silicon Valley, I can confirm that the education is absolutely world-class. Although it’s hard to read that lower-income individuals who have access to the same educational systems experience such different outcomes from their peers in better socioeconomic situations, I do think it’s really cool that we lead the nation in economic mobility. Still, these national numbers are woefully insufficient and they only begin to scratch the surface of the economic mobility problem.
Once someone has completed (or abruptly ended in many cases) the educational portion of their life, the vast majority will enter a socioeconomic class and be stuck there long-term. Because of barriers to entry in high-paying professions, along with the stigma around high-earning trades in America, it is tremendously difficult for an adult with low income to achieve economic mobility for their family within their own generation. In an inevitable future where automation is, in the words of venture capitalist Marc Andreessen, “eating the world,” these issues with economic mobility are only going to get worse unless we dramatically redefine “education.”
In a future which is ruled by automation and rapid innovation, we will constantly need to develop new skills and add additional skills to stay ahead of the growing capacity of AI/manual labor automation, moving “upstream” to jobs that still exist. We’re woefully unprepared to re-educate the world’s truckers and factory workers to prepare them for new career paths.
The gold standard of re-skilling in 2021 is Lambda School. Lambda School is a venture-backed startup focused on teaching coding skills to adults who are looking to switch careers. Lambda’s Founder/CEO, Austen Allred, has a backstory that’s nothing short of incredible. He learned to code while living out of his car in Palo Alto, committing himself to becoming part of Silicon Valley. I highly suggest this podcast where he talks about his experiences.
He also has some really interesting thoughts on the future of work and economic mobility. In short, Lambda is led by a guy who has absolutely “walked the walk” of economic mobility and taken a fairly unconventional path to a career in tech.
Austen’s insight with Lambda was twofold. First, he saw that insufficient resources existed to help people learn to code and he could build a better system, partially based on his experiences of learning outside the traditional academic system. Second, he saw that, in order to maximize economic mobility for his students, he needed to make the program free. To attract and support the type of students who most needed Lambda School, he needed to find a way to fund the program without requiring upfront payments from his students. The solution was the Income Sharing Agreement (ISA) and I’ll get into that more later. The other interesting thing about Lambda is that it does more than isolate the “education” part of the traditional undergraduate experience. Austen and Lambda felt that career services in the traditional educational model are woefully insufficient, and he decided that Lambda should spend an inordinate amount of time helping his students find great jobs. In that sense, he’s been wildly successful - it feels like he’s tweeting every few days about a student who went from earning $40k in their past job to earning six figures in an incredible new job in software development, with a path to earning more than they ever could’ve dreamt. Lambda is a pretty incredible program doing awesome work, but it’s very much just the beginning. It’s a very hard program to scale - the nature of the education and the career services is hands-on. Lambda may ultimately capture a large portion of the overall re-skilling market but, more than anything, they’re creating an entirely new category of EdTech startup.
“Learning to code” is a great place to start with ISA-based adult re-skilling programs, but the same idea can translate to a ton of jobs. We’re a long way away from robots doing the job of a plumber or an electrician, and those are steady, high-paying jobs. Because of the stigma against trade schools in modern society, kids everywhere look down on trades like these, and they’re seriously under-staffed. Trade schools are a great, under-explored option for young people, but they aren’t as feasible for adults looking to make a career transition. They can’t afford the loss of income, and the current models of trade school aren’t necessarily set up for displaced truck drivers or factory workers. This is where the ISA model has massive potential to transform adult education and the future of career paths as we know them.
Income Sharing Agreements (ISAs)
Someone with low income and minimal savings in the middle of their life can’t afford to finance their re-skilling education so they can become viable candidates for new-age jobs, let alone take time off from their current jobs to pursue something new. Lambda School and other startups like Lambda which are disrupting the traditional education industry and pioneering the nascent re-skilling industry have adopted a solution to this problem (and potentially a solution for the student debt crisis) - income sharing agreements. In an ISA, a student does not pay for their education up front, it is financed by the school (more accurately, the ISA is securitized and sold to investors.) Students pay their tuition as a percentage of their income for the first few years after graduation, capped at a reasonable maximum payment (Though I’m not sure the cap is required, or even a particularly good idea.) The great thing about income sharing agreements is that, if you don’t get a relevant job after the program, you don’t have to pay for anything! The school takes on 100% of the downside risk (you risk wasting your time, but 100% of the financial downside) and you share in the upside (but you get most of it.) It’s a fantastic trade for students, a great fix for student debt issues, and the incentive alignment is fantastic. Because the school takes on the downside risk in a way that traditional tuition models don’t facilitate at all, the school has to put together a fantastic product with excellent career services and respectable credentials or they won’t be a viable institution. Incentive alignment is incredibly powerful and ISAs are likely a huge part of the driving force behind the powerful future of education.
I have no interest in replacing traditional college with a pure professional development + career services program - I love college! But I do think ISA-based incentive alignment would have a tremendously positive impact on the quality of career services and career-minded decision making in colleges and their administrations around the country.
Lasting Effects of Covid
Following the Covid-19 pandemic, I strongly expect that students will re-evaluate the value proposition of traditional colleges. Surely, many of these students (myself included) will conclude that college is an incredible, unique experience and they are happy to pay for the bundle, with all its flaws. I am also sure that many students will not feel this way, and that changing tides will threaten the viability of lower-end institutions that have already been struggling for years. Legendary venture capitalist Don Valentine liked to ask every entrepreneur the same question, one that Sequoia (his also-legendary firm) feels is pivotal to predicting the potential success/demise of a promising startup: “Why now?” Timing is (nearly) everything and the Covid-19 pandemic, with some help from rapidly-developing remote work/education tools, has opened up fantastic opportunities for disruptive edtech companies. I am excited to follow the path toward a brighter educational future that will develop during the 2020s.
I often say that the future is bright in almost every category and every sense of the phrase, and the future of education is as bright as any. I am ready to embrace imperfect, unbundled educational experiences on the journey to creating the education of tomorrow.