Why “Learn to Code” Failed

Summary

Between 2011 and 2021, there was a surge in the number of Computer Science graduates, notably at UC Berkeley, due to a societal push for coding skills. However, this boom led to oversaturation, with more graduates than available jobs, contradicting the initial 'Learn to Code' mantra's promise of lucrative tech careers. Universities and bootcamps struggled to meet the demand, leading to strained resources and disillusioned students. As the tech industry faced downturns and layoffs, the oversupply of programmers made workers expendable. This highlights the necessity of adaptable, non-industry-specific skills for future employment stability.

Highlights

• From 2011 to 2021, UC Berkeley saw over 1000% increase in CS graduates, creating more supply than demand 🎓.
• Tech leaders and politicians promoted coding as 'the future,' leading to increased, but often misguided, interest 👔.
• Many students left bootcamps with debt and inadequate job prospects, despite promises of high-paying jobs 💼.
• Universities had to turn away students due to overwhelming demand for CS programs, demonstrating infrastructure strain 🎒.
• The 'Learn to Code' mantra falsely promised job security, revealing the importance of diverse skills over specific coding knowledge 🛠️.

Key Takeaways

• The push for everyone to 'Learn to Code' led to an oversupply of programmers, exceeding the demand in the job market 🚀.
• Universities couldn't keep up with the demand for computer science education, leading to overcrowded classes and underprepared graduates 📚.
• Coding bootcamps aimed to fast-track careers, but many faced similar challenges as universities, including high costs and limited effectiveness 💸.
• The tech industry's volatility exposed the fallacy that coding was a stable career choice for everyone ⚖️.
• The true value lies in adaptable skills and problem-solving, not just learning to code 💡.

Overview

Once heralded as the golden ticket to high-paying jobs, the 'Learn to Code' movement flooded the tech industry with a wave of new programmers. However, the demand for these skills quickly outpaced the job market's ability to absorb them, leading to an educational bottleneck and numerous graduates finding themselves without the promised career opportunities.

Universities and coding bootcamps both tried to accommodate this surge by expanding programs and promising swift, lucrative tech careers. Yet, they often overpromised and underdelivered, with overcrowded classes and hurried curricula that left many with skills misaligned with industry needs. High expectations met hard realities, leaving many students disillusioned.

Amidst this, the tech industry faced a fluctuating market that underscored the inherent risk in betting on any single skillset. The narrative that coding could universally solve economic challenges proved overly simplistic. The lesson? True career resilience lies not in narrow expertise but in cultivating a broad set of adaptable, problem-solving abilities.

Chapters

• 00:00 - 02:30: Introduction and Contextualization of "Learn to Code" The chapter discusses the striking growth in Computer Science graduates at UC Berkeley, highlighting a 1106% increase over a decade. It humorously notes that if this trend continues, all undergraduate students will be studying Computer Science in seven years, and by 2059, the school might produce more 'Computer Scientists' than California's population. It also touches on the current employment status of software developers.
• 02:30 - 06:30: Rise in Popularity of Coding Education The chapter discusses the significant increase in the popularity of coding education, particularly highlighting the dramatic rise in interest in computer science programs at universities across the United States. It mentions specific institutions, such as Berkeley, CalTech, and MIT, that have become epicenters for computer science education. Despite Berkeley's reputation, it is noted that even more significant growth and interest are observed at other institutions. The transcript provides a snapshot of how these programs have gained traction over the past six years, indicating a major shift in academic and professional focus towards computer science.
• 06:30 - 10:00: Challenges Faced by Universities in Meeting Coding Demand The chapter explores the challenges that universities face in meeting the increasing demand for coding and computer science professionals. It highlights a disparity in graduates from different fields, noting that while only seven students graduated with a Bachelor of Science in Chemistry from MIT, two-hundred sixty-six students majored in Computer Science and Engineering. Additionally, the largest major at MIT, Electrical Engineering and Computer Science, is more than twice as popular as the Computer Science and Engineering major.
• 10:00 - 16:00: Introduction of Coding Bootcamps The chapter discusses the transformation of U.S. colleges over the past 15 years into more specialized institutions focusing on technology and computing. It highlights the development of dedicated colleges for computing, such as those at Berkeley, MIT, and Cornell, pointing out a significant shift in educational focus.
• 16:00 - 23:00: Challenges Faced by Coding Bootcamps The chapter discusses the emergence of tech startups with the release of the iPhone in 2007, followed by platforms like Uber, Airbnb, and Instagram. It highlights the cultural shift in America from traditional Wall Street to dynamic Bay Area startups, emphasizing how shows like 'The Social Network' and 'Silicon Valley' romanticize tech founders as modern-day, meritocratic visionaries compared to historical industrial tycoons.
• 23:00 - 27:00: Impact of Economic Downturn on Tech Employment In this chapter titled 'Impact of Economic Downturn on Tech Employment', the discussion highlights the significant endorsement of programming and coding by national leaders, elevating it to a pivotal future career path. Former President Obama is noted for describing coding as a 'ticket to the middle class' and for promoting initiatives like Computer Science Education Week and 'Hour of Code'. He also warned of an impending shortfall of 1 million STEM graduates by the year 2022, underscoring the importance of such skills amid economic challenges.
• 27:00 - 34:30: Limitations and Criticism of "Learn to Code" The chapter discusses the limitations and criticisms of the 'Learn to Code' initiative. It mentions the appeal of the slogan 'Learn to code' to various groups, such as New York Mayor Michael Bloomberg, who in 2012 announced his intention to learn coding, and other politicians and corporations. For Republicans, it aligned well with their focus on vocational and career-oriented education, while corporations saw it as a way to ensure a steady supply of skilled labor.
• 34:30 - 39:30: Conclusion and the Value of Adaptable Skills The chapter discusses the rationale and impact of promoting coding skills across different societal and political perspectives. Coding is seen as a key asset for national security, economic competition, and expanding opportunities. It highlights the bipartisan appeal of coding as a tool for empowerment during economic recovery.

Why “Learn to Code” Failed Transcription

• 00:00 - 00:30 Between 2011 and 21, the number of Computer  Science graduates at UC Berkeley increased   by one thousand, one hundred, and six percent. If this trend continues — as one  professor observes — all 30,000   of its undergrads will study Computer  Science in just seven years. By 2059,   the school will churn out more “Computer  Scientists” than there are people in California! Meanwhile, get this: there are fewer  software developers employed in the
• 00:30 - 01:00 United States today than there were six years ago. Clearly, something has gone terribly wrong. And Berkeley isn’t even the most Computer  Science-obsessed university. Not even close. This is UCB, this is CalTech, and this is MIT. You’re not reading this wrong. A full forty  percent of the school that once helped invent
• 01:00 - 01:30 the radar, spreadsheet, and lithium-ion  battery studies the same one subject. Last year, a grand total of seven students  graduated from MIT with a Bachelor of Science   in Chemistry. Two-hundred and sixty-six majored in  Computer Science and Engineering. And that’s not   even its largest CS major — Electrical Engineering  and Computer Science is over twice as popular.
• 01:30 - 02:00 Over the past 15 years, U.S.  colleges have transformed from   multidisciplinary institutions  of higher education to Microsoft   training programs that also dabble in  philosophy and physics on the side. In 2023, Berkeley joined MIT and  Cornell in even creating an entire   College of Computing — its first new  school since Eisenhower was president. It’s not hard to see how we ended up here.
• 02:00 - 02:30 The iPhone was released in 2007; Uber, AirBnB,   and Instagram soon after. America’s cultural  center of gravity was shifting from stuffy   Wall Street board rooms to youthful Bay Area  garages. Shows like The Social Network and   Silicon Valley mythologized the tech founder  — the 21st century’s more meritocratic,   more visionary, and certainly more eccentric  answer to the industrial tycoons of the past.
• 02:30 - 03:00 Then, our nation’s leaders endorsed  programming with their official   seal of approval, elevating it from a  promising new career to “the future.” While in office, President Obama  called coding a “ticket to the middle   class.” The White House celebrated  Computer Science Education Week,   promoted “Hour of Code,” and warned of a dire  shortfall of 1 million STEM graduates by 2022.
• 03:00 - 03:30 New York Mayor Michael Bloomberg  even announced he, personally,   would “learn to code” in 2012 (although  there’s no evidence he ever did). There was something for everyone in  that 3-word-mantra, “Learn to code…” For Republicans, this directive  neatly aligned with their vision   for more vocational, career-oriented education. For corporations, it promised a steady  supply of skilled labor — labor — I might
• 03:30 - 04:00 add — that came pre-trained, at taxpayer expense. For national security hawks, coding  helped strengthen America’s global power,   enabling it to “out-compete” China. If diversity was more your cup of tea,  one could tell a pretty convincing   story about how this was really  all about expanding opportunity. And for Democrats steering the  economy through the Great Recession,   “learn to code” sounded empowering, hopeful,  and — most importantly — vague enough to absorb
• 04:00 - 04:30 virtually any of the difficult questions posed  by globalization at an especially fraught moment. Layoffs? Automation? Outsourcing? …Not to worry if we all just rolled   up our sleeves and, say it  with me: “learned to code!” Red and blue states alike rushed to teach Python  and Javascript in classrooms. Today, nearly   15,000 U.S. high schools offer Computer Science  classes. Even 37% of middle schools do. Eleven
• 04:30 - 05:00 states go even further: requiring all students  take a computer science class to graduate. Absent any dissenting voices, there was  every incentive to exaggerate the ease of   coding and romanticize the life of a programmer. Obama told us “just about anyone can become” a  computer scientist, quote, “with a little hard   work…” Years later, President Biden echoed  this sentiment in his signature no-nonsense
• 05:00 - 05:30 tone: quote “anybody who can throw coal into a  furnace can learn how to program, for god’s sake.” One popular article from 2012 — when U.S.  unemployment was still well over 8% — declared,   “to anyone out there who says you can’t  get a job: You can have one. A fun one.   Learning code is not about numbers and  mathematics. It’s more like architecture…” Well, young Americans didn’t  need to be told twice!
• 05:30 - 06:00 Younger Millennials and older Gen Z looked  around, saw their siblings struggle to find   work after the Great Recession, and  decided they’d take the ‘cool’ job   that lets you earn six figures working from  a bean bag while wearing a hoodie, requires   a bachelor’s degree at most, and is apparently  remarkably “quick” and “easy” to learn, please. But although Microsoft and  Facebook and Apple were ready   for this massive influx of programmers,  universities most certainly were not.
• 06:00 - 06:30 The popularity of college majors, of course,   fluctuates all the time. Seasoned law school  admissions officers still remember the sudden   rise in applications following the hit  80s drama “L.A. Law,” just as pharmacy   school briefly became popular during  the early 2000s pharmacist shortage. But the Computer Science boom  was something else entirely. The basic problem for schools was this:
• 06:30 - 07:00 while the total number of undergraduate  CS students tripled in just 10 years,   the number of Ph.D. students — who become  professors — has stayed more-or-less the same. The reason for this is no mystery: a Ph.D. student  can expect to receive a $40,000/year stipend — if   they’re lucky. That same student can easily  earn $200,000 or more at Amazon or Netflix. In other words, there isn’t anyone  to teach all those undergraduates.
• 07:00 - 07:30 The result, in many Computer Science departments,  is an impersonal, factory-like experience.   Professors are perpetually stressed and invariably  overworked. Class sizes are massive — 4, 5,   6 hundred students in introductory courses.  Undergrads are even deputized as TAs — meaning   the person who took this class last semester  might be the only person you can ask questions.
• 07:30 - 08:00 And yet schools still turn eager students away. It’s become common at many universities to  make students re-apply to their Computer   Science program at the end of their  first or second year, for example. At Swarthmore, the University of Maryland,  and UCSD, students are entered into a lottery!   Meaning, you can apply and be accepted to  Swarthmore (not to mention pay just over   $90,000/year) only to later be denied access  to your preferred major by random number.
• 08:00 - 08:30 Needless to say, this leaves many students  feeling… disillusioned with the college   experience. They graduate with debt, with little  career help from their school, whose resources   are spread thin, and with minimal — if any —  contact with their professors along the way. On top of all that, many of these students  feel they were inadequately prepared for the   job market. Because many CS programs have their  origins in math and engineering departments,   their curricula often focus more on the how  — the general algorithms, ways of thinking,
• 08:30 - 09:00 and processes — than the what — the specific  syntax, programming languages that are currently   marketable, or skills tested in modern job  interviews. Graduates will often have dabbled   in half a dozen languages but truly mastered  none, forcing them to learn on their own. It wasn’t hard to imagine  there might be a better way. And this being the era of quote “disrupting”  “outdated” industries like taxis, hotels,
• 09:00 - 09:30 and DVD stores, the proposed  “solution” was wrapped in lofty,   emancipatory language: education was about to  be revolutionized; universities, out-innovated. Enter: the coding bootcamp. The idea was simple: if you already  knew what you wanted — a six figure   job at Amazon — you could forgo the 4-year  Computer Science degree, pay just 10 or 20,   or 30 thousand dollars, and “learn” to code in a  fraction of the time — usually 12 weeks or less.
• 09:30 - 10:00 By stripping out general  education requirements and   teaching to the test (tech interviews  often ask the same set of questions),   they promised a streamlined, more efficient,  and more accessible route to the promised land. At their peak, hundreds of bootcamps  graduated nearly half as many students   as 4-year Computer Science programs and  took in north of $200 million a year.
• 10:00 - 10:30 But as they grew, as they began appearing on  billboards and in TV commercials and attracting   a wider and wider group of students, they began  feeling many of the same strains as universities. First, they realized their students  were lending at least as much of   their reputation to them as they were to students. In other words, if one of them failed an  interview or was fired, that employer or   recruiter would remember and associate their  performance with the bootcamp on their resume,
• 10:30 - 11:00 making it harder for future students to  land jobs there. Silicon Valley, after all,   is a small world. Most students are looking  for work from one of just five companies. Thus, they began carefully  guarding their reputations,   selecting only those applicants who  enhanced their brand. One bootcamp,   Hack Reactor, for instance, has an acceptance  rate of just 3% — about the same as Harvard.
• 11:00 - 11:30 Another problem was a lack of teachers. Also  like universities, bootcamps struggled to   compete with the salaries on offer by Big  Tech, often resorting to hiring their own   recent graduates who couldn’t find work elsewhere,  inflating their employment figures in the process. Finally, they faced pressure to raise tuition. Initially, bootcamps pitched themselves  to investors much like WeWork did:   as software companies. They could  develop their curriculum once and
• 11:30 - 12:00 then churn through students  with zero marginal cost. But, like WeWork, reality  turned out to be more complex. In practice, bootcamps are either little more than  classrooms filled with textbooks — in which case,   they don’t offer much that those same  textbooks alone can’t for a fraction of   the cost — or they’re highly-structured,  personable, guided experiences in small,   intimate settings — in which case, that’s  expensive! As the number of students grows, you’ll
• 12:00 - 12:30 need to keep hiring new teachers, by definition,  to maintain the same student-to-teacher ratio. If anything, the marginal cost went up. Later  students, attracted by their increasingly “get   rich quick-syle” marketing, had less familiarity  with and less interest in programming and   therefore required more guidance than their  earlier, more intrinsically motivated peers. Just to maintain the same outcomes, therefore,
• 12:30 - 13:00 bootcamps had to raise prices. And that’s a  problem when students are paying out of pocket! Bootcamps wanted in on those  sweet, taxpayer-funded federal   student loans. But, as unaccredited  institutions, they weren’t eligible. Meanwhile, for reasons we’ve  covered in previous videos,   colleges found themselves short on cash.  They wanted — needed — a slice of that sweet   bootcamp revenue. But they aren’t very good  at recruiting non-traditional students. And
• 13:00 - 13:30 the Higher Education Act of 1965 prevented  them from paying third-parties for bringing   them students after those incentives led  to aggressive and misleading marketing. …Except that in 2011, the Department  of Education created a loophole:   schools could pay third parties to bring  them students if those payments were part   of a larger “bundle of services,” of  which recruiting was only one part.
• 13:30 - 14:00 In other words, as long as a school didn’t call   these payments “kickbacks,”  they now had a green light. A university would “partner” with what they called  an OPM — an “Online Program Manager” — this was   the bootcamp. The OPM would create the curriculum,  hire the teachers, and recruit the students,   which the university would then slap  its name on, usually for a 40% cut.
• 14:00 - 14:30 As far as the government was concerned, these  programs were accredited, so bootcamps got   access to student loans. Students saw a  trusted 100-year-old brand name, not some   random fly-by-night bootcamp. And universities  could get away with what was otherwise illegal. Put differently: bootcamps became  more and more like the “unwieldy,”   “inefficient” 4-year universities  they originally sought to “disrupt,”   until they were swallowed whole  by those very same universities.
• 14:30 - 15:00 Now, this more-or-less worked through  the pandemic, when growth in the tech   industry was off the charts and companies  raced to hoard as much labor as possible. In one 2023 Wall Street Journal  profile, a woman said she was paid   $190,000/year by Meta to do… almost nothing  — presumably because aggressive hiring at   the time was rewarded by investors,  and competition for talent fierce.
• 15:00 - 15:30 Then came the inevitable crash. Nearly half a million tech workers  were laid off in 2023 alone. Another   quarter million lost their jobs  once in 2022 and again in 2024. In total, that’s roughly the  number of U.S. manufacturing   jobs lost from the early 2000s “China Shock.” The unemployment rate in tech is now  higher than the national average.   Recent graduates have had their  offers rescinded. And many professors,
• 15:30 - 16:00 including this one at Berkeley —  one of the most well-known Computer   Science programs in the nation — report that  even their best students can’t find jobs. Bootcamps are doing even worse — if they  still exist at all. One, called “2U,” declared   bankruptcy last year. Launch Academy announced  it would “pause enrollment.” And Dev Bootcamp,   which was acquired by education giant  Kaplan, closed its doors forever.
• 16:00 - 16:30 It all happened remarkably fast. It seems like only yesterday that tech  was “the future.” Not so long ago,   the White House predicted a shortfall  of 1 million STEM workers by 2022. Instead, there are fewer  software developers in the   United States today than there were six years ago. How did this happen? Well, it’s not rocket science: “Learn to code,”  meet another 3-word mantra: “supply and demand.”
• 16:30 - 17:00 Recall that in just ten years, the  number of Computer Science grads at   Berkeley increased by over a thousand  percent. This was always unsustainable. Now, to be clear: “Learn to code”  didn’t cause this recent downturn. In 2022, U.S. interest rates rose to their highest  level in 15 years — from nearly 0% to well over 5.
• 17:00 - 17:30 And the tech industry is unusually exposed to this  number — making software requires massive up-front   investment. When the cost of borrowing goes up,  making software becomes a lot more expensive. But there’s no doubt that the over-supply of  programmers — driven by “Learn to Code” — made   workers more expendable in the eyes of employers.  When programmers are so abundant, it’s a whole   lot easier to fire them on demand — treating  them as a spigot they can quickly turn on when
• 17:30 - 18:00 needed and off when not, rather than a resource  to invest in and retain through thick and thin. Every industry, of course, has ups and  downs. And brutal as this downturn has been,   growth will eventually return — the Bureau of  Labor Statistics still expects computer-related   occupations to grow 11% by 2033. The same,  to be fair, cannot be said of all fields.
• 18:00 - 18:30 Still, a starry-eyed 18-year-old  back in 2016 could be forgiven   for thinking that tech was somehow… different. For years, common sense had been  thoroughly drowned out by the   increasingly religious zeal of “Learn to code.” “Learn to code” was more  than a gentle encouragement   to consider programming as a potential career.   It was an all-encompassing vision of “the  future,” unbounded by the laws of economics.
• 18:30 - 19:00 That 2012 article from earlier, for example,   compared coding to reading and writing — a  surprisingly common sentiment at the time.   The message wasn’t that programming  was a useful skill, it was that soon,   everyone would need to program. Otherwise  serious people argued that everyone from   electricians to teachers to insurance adjusters  would be coding in the not-so-distant future.
• 19:00 - 19:30 If you accept that premise,   one thousand percent growth in Computer  Science graduates doesn’t sound so crazy. But if programming is not a universal  skill like reading or writing,   if it’s still just one, admittedly  often lucrative career among many,   then it just means a thousand percent more  competition for a limited number of jobs. Today, there are something like 1.9 million people  employed in tech. Even if we double that estimate,
• 19:30 - 20:00 it would still only represent  2.3% of the U.S. labor force.   Programming simply could never have  absorbed all seven million unemployed   Americans. Never mind that someone still needs  to manufacture the keyboard you program with,   or feed the person who manufactures that  keyboard, or repair that person’s internet. Tech is also at least as  volatile as any other industry. The sudden and stratospheric  rise of AI is a prime example.
• 20:00 - 20:30 These ups and downs become  clear when we zoom back in time. This recent boom has just been so  unusually long that, until recently,   many young people had never  experienced anything else. But “learn to code” wasn’t just at odds  with Econ 101, it was also ignorant of,   or at least indifferent to, the  wide diversity of human interests,   talents, personalities, and life circumstances.
• 20:30 - 21:00 Even the most accomplished programmer will  tell you: building software can be quite   difficult. Like any skill, not everyone will  be great at it. Nor, for that matter: enjoy   it — for the same reasons not everyone enjoys  working in an office, or sitting down all day,   or staring at a computer, or working on a team,  or solving complex problems on a deadline. This is all quite obvious when you  replace “code” with “repair medical   equipment,” “study epidemiology,”  “become a nurse,” or any other job.
• 21:00 - 21:30 Overlooking the wide spectrum of human skills,   interests, and circumstances was convenient. And  although it sounded empowering as a soundbite,   it was often much closer to  exploitative, in practice. “Learn to code” reduced people to  interchangeable units of labor. Coding bootcamps represented the most distilled  version of this one-size-fits-all “solution”:
• 21:30 - 22:00 promising that single mothers,  recently laid-off factory workers,   and well-heeled highly-computer-literate  18 year-olds could all be funneled into   a bootcamp and emerge, 12 weeks later, as  Computer Scientists making $150,000/year. Instead, many single parents working two  jobs earnestly enrolled in bootcamps only   to discover they offered minimal support for and  couldn’t accommodate non-traditional students. Former general contractors struggled to keep up  with their peers who had prior coding experience.
• 22:00 - 22:30 And since they were told learning to code is  “so incredibly easy,” they blamed themselves. Thousands graduated with nothing more  than a $20,000 hole in their bank account,   3 months of lost wages, and  a line on their resume that,   in the eyes of many employers,  worked against, not for them. Others, of course, were luckier. But, perhaps most humbling of all, even many of  them ultimately fell victim to “Learn to Code.”
• 22:30 - 23:00 They did exactly as they were told, set their true  passions aside and pursued coding as a career,   were admitted into a selective bootcamp  or Computer Science program, studied hard,   were fortunate enough to land a competitive,  highly-paid job at Amazon, and after all that,   were still one of its, say, 18,000 employees laid  off with the stroke of a pen on January 5th, 2023.
• 23:00 - 23:30 No matter how many times it's repeated, no  matter how many high schools it’s taught in,   “learn to code” is no more a magic solution to  economic uncertainty or job insecurity than “learn   to service wind turbines” or “learn occupational  therapy” — demand for which, by the way,   is expected to grow by 60 and 22%, respectively  — far faster than software developers. Everyone, everywhere is at the mercy of  the labor market. The only solution — the
• 23:30 - 24:00 best we can do — is be adaptable and invest in  non-industry-specific skills like problem solving. Today’s sponsor, Brilliant, can help you do just  that — spread your eggs across many baskets. Brilliant is the perfect sponsor  for this video because they believe,   as I do, that what’s truly  valuable in this fast-moving   era we live in are higher-level skills like  critical and strategic thinking, creativity,   and problem solving. Programming languages come  and go. We forget the formulas and facts we read
• 24:00 - 24:30 years ago in a textbook. But the way you approach  and solve problems stays with you forever. Brilliant takes that approach  and applies it to logic, puzzles,   data analysis, and geometry. They have courses  on “Real-World Algebra,” “Thinking in Code,”   “Scientific Thinking,” “Everyday Math,” “How  AI works,” “Geometric thinking,” and much more.
• 24:30 - 25:00 Sure, you’ll learn to program in  Python, to calculate probability,   and analyze data. But you’ll also sharpen  your mind, learn to think like a programmer,   scientist, or mathematician, and develop  a powerful learning habit along the way. Ironically, by focusing a little more on the  process and a little less on the outcome,   you actually get better results: by  making learning fun and interactive,   rather than all about “getting  the right answer,” Brilliant’s
• 25:00 - 25:30 beautifully designed lessons make you  want to keep coming back for more. To check out Brilliant completely free  for 30 days, click the link on screen   now or in the description below —  that’s Brilliant.org/PolyMatter.   Doing so will also get you 20% off  an annual premium subscription. Go, pick what interests you  most, and start learning today!