Discover more from Zayn’s Newsletter
Tech companies will build the future of education
Why and how big tech will build their own education curriculum, programming, and on-job experiences.
I’ve structured this post into two parts:
Bullet point update on 2022; the last time you heard from me was December 2021
Essay on the future of education, as I see it, transitioning from traditional schools/universities to corporations/big technology (note: the entire essay is a 16-minute read. because of its length, it was truncated by substack’s email distribution. to read the complete essay, without interruption, click here)
Bullet point update on 22’
Finished my first semester @Babson College and wrote a memo to the president about how Babson can recreate itself to be a magnet for technical founders. 2
Read 18 books.3
Essay on education transitioning from schools to corps
The origin of this idea, education transitioning from high schools/universities to corporations, came from a stat I mentioned at Collision:
52% of high school students are prepared for college. 66% of college students are prepared for a job. As students complete more education, they feel less prepared for the next phase; an idea counterintuitive to the concept of education. I think the high percentage of unprepared students is created by governmental agencies setting the curriculum instead of industries, which know what students need to know for a job. Education for software engineers or computer scientists should be set by companies like Google, MSFT, Apple, and Amazon, the best technology companies in the world.
You can view the clip here.
The point is that a company like Google should create a curriculum for software engineers instead of universities because the companies understand how their skill needs to change. In contrast, the universities are getting this data, usually at a lag, and statistically, haven’t been able to thoroughly prepare students for the job.
As I continued my research on the skills mismatch created as students progress through school, it became more apparent that everyone (corporations, schools, students) recognized it as a problem but didn’t change anything. Curriculums and budgets stayed the same while the mismatch grew.
Data on the skills mismatch between college and job
A report by BCG found tremendous skills mismatch in the global workforce responsible for a sizable loss in labor productivity; researchers estimated that 1.3 billion people worldwide have competencies misaligned with their work, including 53.3 million in the US. In other words, the skills students learn in college aren’t helping them prepare for the workforce, exploiting the misalignment between college curriculums and workforce needs that are changing at a non-linear pace.4
A recent article from Paragon One mentions that pre-skilling, providing employees with the skills they require before they begin their career is what higher education institutions were created to do. 36% of current company leaders believe this is what institutions are doing.5
A survey of 2,000 students showed that they favor high school degrees that focus 75% of the curriculum on skills related to the real world and 25% on skills related to higher education. But, in some high schools, the current focus of curriculums is 100% skills related to higher education and 0% associated with the real world. This is brilliantly illustrated in the North Thurston public school district budget. While regular instruction expenditure increases from $135M to $150M over the next four years, skill center instruction6 remains at $0 during the same time.
Apple CEO Tim Cook said there are in-demand skills students aren’t learning in college. In response, Apple piloted an ‘everyone can code program’ in the United States, and 4,000 schools are already using it. Apple’s school involvement has allowed them to recruit and hire more people without college degrees6. In other words, the best in the world are creating programs for students, and it's working. It’s also a clever way for Apple to get more students in schools to begin developing in swiftui and achieve certifications in this language, increasing the talent pool of swift developers that Apple can hire. The program includes free resources for teachers to follow along with students, but teachers don’t need to know coding7.
Everyone Can Code as a case study for ed’s future
I think the ‘everyone can code program’ foreshadows what Apple’s future in education could look like. They started by teaching Kindergarten students how to code. Still, Apple also has hardware designers, product managers, copywriters, corporate lawyers/general counsels, etc., giving them a range of skills they can teach students. For example, like software engineering, I envision Apple designing a curriculum that works up from the fundamentals, increasing the difficulty and number of skills as students progress in school. These skills include learning how to code in an object-oriented programming language, mastery of calculus, communicating with project team members, etc.
As each student builds their knowledge base, they’re also building projects using Python, SwiftUI, Java, or Apple APIs and uploading them to Apple’s Developer project showcase. When high school internships open, Apple can search for an engineer with mobile/web development experience, a strong basis in math, and project manager experiences/results, and as they scope down, review a list of N individuals and select the engineer that matches the skills/competency they were looking for. As part of your project building, you were required to get one/two projects approved by an engineering executive at Apple. When a manager reviews your portfolio, they see this and note it. Each company creates its version of this, replacing LinkedIn as it is and creating a new talent system that combines a high-quality curriculum built by the company with a project database flowing with uploads from students.
Four high-quality internships at multiple software companies, a robust portfolio, and the essential software and computer science knowledge built through Apple’s curriculum eliminate the need for a college degree.
And this idea isn’t limited to software. Tesla can create a curriculum to teach aspiring students of robotics and automation how to build factory-based automation systems. Ginkgo Bioworks can teach students about DNA replication. Varda Space Industries can teach students about building manufacturing in space.
Delian at Founders Fund mentioned that two people at Founders Fund understand the space industry well. Because of this scarcity, the firm is less likely to invest in this area. But to learn more about Earth and the planets around us, firms like Founders Fund must invest millions of dollars in space to find a couple of winners. Part of the problem is that the educational infrastructure to teach people about areas doesn’t exist as it does for software engineering. There are more than 150 coding bootcamps and 0 space boot camps. A company like Varda, looking for more aerospace or space engineers, could create a curriculum to teach someone about space, reducing the scarcity of people working in this industry.
Companies are incentivized to do this. Big tech companies spend $2.5B on retraining employees yearly because of the skills mismatch from college to the job. In addition, each new hire costs $4,000, and if you take the total number of employees to google hired in Q2 of 2022, 10,000, the total cost is $40M. Assuming this spend stays the same, that’s $160M or 25% of google x’s moonshot factory budget. In other words, if Google had a lower cost of hiring employees through a pipeline of early and recurring recruits, they could spend more money on long-term, world-changing projects like Waymo or Mineral.
The curriculum/what to teach students and experience so that they can apply the theory are two core tenants of education. Unconventional wisdom says you don’t need a school for either of those. Interestingly, as I was redrafting this essay, a16z posted crypto courses to their youtube channel. Someone will aggregate this content into an a16z education platform where users are completing projects, backlinking to the learning they did, creating a forum or showcase for those projects, meeting a co-founder, and raising with a16z. Education projects like these remove the gates that traditional schools are built with and allow anyone with internet access and a desire to learn to make the next big company.
Data on middle-class education (enrollment and diversity)
I think companies can use the same elementary-high school programming and curriculums to create the qualification infrastructure for the middle class. Here’s some data on what’s being done:
Middle-class enrollment in college has declined by 5% since 2010. Suppose big tech companies took over the curriculum at the elementary-high school level. In that case, you could give students enough training (comp. approved) and job experience to prepare them to join one of those companies without needing to enter college. Some in the middle class enroll in online tech courses to complete certifications and get hired at companies paying 2-3x higher than their previous job. While the pandemic could partly skew the hiring numbers, the pattern of new companies like Okta prioritizing skills and experience over college has helped diversify the workforce and recruit people who usually would’ve had access to these jobs, locked by the college gates because they couldn’t afford the tuition. This is one of the key ideas behind increasing company dollars and education courses; more people who avoid college because of the threat of debt can obtain job stability and positions through these certifications and on-job experience. And it closes the gap between what companies want and the skills their recruits bring.8
Google has had early success with its career certificate program9 on hiring, diversity, and salary axes. They've started to create an additional pathway for students who don’t have a degree.
Hiring: Deloitte has hired 14 program certificate holders since 2021.10
Diversity: When you place a bachelor’s degree requirement on a job, you eliminate 70% of African Americans, 80% of Latino workers, and 70% of Rural Americans. More than half of the career certificate program participants are from the underrepresented groups above.
Salary: Almost half of the participants in the certificate program come from the lowest income tier, reporting annual wages of less than $30,000 when they enroll in the program. For some jobs, the average salary of jobs available to these holders after they complete the program is $63,000 and higher.11
The Bloom Institute of Technology (formerly lambda school) makes the non-college path more lucrative. In the next 5-10 years, they’ll create an education pathway that makes aspiring software engineers look irresponsible if they don’t take this path. Their CEO, Austin Allred, recently tweeted out that out of their current cohort of backend developers, 22 devs have landed roles before graduation, 90% of these pay more than $100k annually, and the students who are graduating with these offers pay $0 until they get hired. My favorite case study is from a high school student who chose bloom tech over college and was recently hired into FAANG.
Micro-internship apps for certificate-completing students.
Career certificates alone aren’t enough. They give students knowledge, not experience. The lack of experience in career certificate holders has the ultimate counterargument to those who have one and haven’t been successful in the job search. I subscribe to the view of Austin Allred; 5+ years of high-quality experience is heavier than an undergrad degree.
Google’s current career certificate strategy has corporate partnerships so students can apply for and receive a job. But these unlock at the end of their certificate. Since students only have a certificate and no experience since obtaining it, the difficulty in getting a job remains a problem. I think Google should change the rules with its existing partnership network of 140+ companies such that certificate graduates who complete a problematic project milestone are offered an internship-like position for two months before they return to achieve the next crossroads where they’re offered another two-month internship. These two internships on the path to completing a certificate are similar to the average of 2+ internships college students hold. The best part about the certificate programs and the future of education broadly is that it’s not about having an A, B, or C in a class; it’s about whether the code runs without bugs, how clean the code is, how much revenue a project brought in post-launch. We can strip away the triviality and fluff of credentials and obsess over the competency, skills, and high standards a student embodies.
Counterarguments to big tech building the future of edu
One counter-argument to this essay's idea is about big tech’s core competency. It’s not education. For Google, it’s search; Facebook, human-scaled connection; Amazon, world-class online retail and technology services; Apple; hardware design, Microsoft, building operating systems and software products. But, I take the view that education is stagnant and largely will be dormant for the next 5-10 years because there isn’t enough reason for college institutions to make changes. Despite record disenrollment, many students are still applying to top colleges, hoping to get a degree; demand is steady if not increasing. I do think that these companies have brilliant people in engineering, design, user experience, and business and can use the knowledge/capability of those individuals to create an education segment that helps bridge the gap between the skill level of graduating students and what corporations need, in addition to my mentions on providing optionality to the middle class.
Another counter-argument to this idea is the lack of college experience, network, and particular credentials. These three come with college degrees, not so much with career certificates. Network: This can be categorized into alums, current, and future students. Alumni networks are deemed crucial for students because they provide graduates working in jobs and trust their alma mater’s curriculum and talent pool enough to return to the university to hire them. It’s a good sell on the university side as it gets students to enroll in universities where they can network their way to a job, not even mentioning the network effect it has for universities. Ex: as they get more CEOs, founders, and bankers who dominate the top ranks of an industry, they recruit more interested students.
I think the alum idea can be supplemented with a mentor-like program, as Keith Rabois described at SaaStr 2019. Out of the 200 students doing a data analyst career certificate at a time (artificial number as an example), you split them up into ten groups, bring in 10 current data analysts at a senior equivalent position that coach the students from the start-finish of their program, weaving in other volunteer data analysts who want to help the students. An obvious question is: can this system be scalable if you have 200,000 students with a data analyst certificate? At a university, the alum network isn’t actively mentoring the students, whereas my above suggestion mentions a medium-touch mentoring system set up by Google. I think in the case of 200,000 students, you’d be able to use the available data (ex: how long a student takes to complete a course, how many times they request help, where they are in their career journey, what is their goal with the certificate) to deploy people accordingly.
I think Facebook’s education structure of live classes, breakout rooms, and active community is interesting on current student networks, but I don’t know if this network is scalable. Offline/online networks where students ask questions in Youtube comments, forums, and discords have shown success in career certificates and equivalent programs. While the richness of the network isn’t available in the certificate programs, I question whether or not the students taking the certification want this network. Future students offer a low-cost yes or no to students pondering the accreditation. Like a university, prospective students reach out to current students and ask about their experiences. How are you enjoying the course, the professors, the relationships you’re making, are you placing in jobs, etc.? I see a similar relationship between current and future students in certificate programs, too.12
Where big tech building the future of ed fails
I think there are two ways this fails:
The first: corporations double-down on upskilling and reskilling instead of increasing their footprint in online education and challenging universities, high schools, etc.
The second: the educational infrastructure to build this is too heavy, and it’s neither intentional nor high standards. The companies that won’t succeed don’t plan for the end2end learner experience (ex: they only deliver the course, not the on-job expertise). Or, they choose courses based on what skills were popular in the past, not timeless or on-demand skills. Additionally, companies will lose if they aim to be just better than universities. Big tech companies know what high standards look like in their own companies, and that model should be applied to education.
Why I’m confident in this idea
Recent news about Amazon launching an online learning platform to offer individuals access to courses in programming and cloud computing at low cost makes me hopeful that more corporations recognize the opportunity to pre-skill, not up-skill students13. Facebook also announced they’re working on a Udemy competitor aiming to offer students live courses. Microsoft is expanding courses in their Microsoft Learn program. And Google recently announced a $100M fund for their career certificates program.
Education is experiencing frustration at every level, from students to parents. A skills mismatch that’s growing, causing corporations to spend more on upskilling/reskilling and a labor shortage of technology positions. These create the conditions for a change in how education is structured. Schools become hubs for students, and corporations become the providers of the curriculum, professors, and classroom resources. We’ve seen many experiences in the past where the public sector has failed, and the private sector has innovated.
Amazon increased the minimum wage in 2019, fading the US government, which keeps the $7.25 minimum wage. Amazon is stopping duopoly and monopoly power from payment networks.
Bezos, Musk, and Branson led the space race instead of NASA.
I think education from companies (private sector) instead of schools (public sector) could be the fifth bullet on this list.
until my next writing,
thanks for reading this post! subscribe to receive new essays.
Beru is on pause now.
I’ll write an essay about this soon. The memo is private now.
My favorite was Invent & Wander , a collection of essays/talks from Jeff Bezos. The book I recommended most was Start-up Nation, a story about Israel’s tech and business ecosystem. The book that made basic concepts visible to me was Joy of X, a book about the profound frequency mathematics has in our lives. If you’re looking for a book like experience in video form, I recommend, The Men Who Built America. It’s a docuseries on the entrepreneurs who re-built America after the Civil War.
Recent LinkedIn data shows that the skillsets for jobs have changed by around 25% since 2015 – and by 2027 that number is expected to double. That means your job is changing, even if you aren’t changing jobs – just as business demands are changing even if you’re not changing your business. (quoted from weforum article)
These students have built and shown Apple they have the necessary skills of the position they applied for. (ex: coding skills if they applied for a software engineering position)
More ‘corporate universities’ are being created to minimize the skill delta between college and corporation.
Positive side effect: gives ~ 4 years back to new employees to learn more about the business they’re working for (ex: someone gets hired by Bank of America, they have 4 more years to learn about banking).
Counter-argument from youtube video: Experience in a job is needed; certification isn’t enough. [I addressed this in the essay.]
These were all data analyst roles.
I don’t understand why tuition isn’t variable based on the major you choose similar to waterloo’s tuition. ex: If you are an art major, you get charged tuition based on projected average earnings over a 20 year career (ideal). The current option is, you pay the same as an engineer who’ll make at a minimum 2-3x more than you over a 20 year career. Interestingly, acceptance rates are variable.
The last and most popular counterargument is about areas like theoretical physics, where the content can’t be learned online or obtained with a certificate like google’s. For these careers, this program doesn’t work, yet.
This is in addition to their career choice program which covers 95% of tuition for programs warehouse workers want to take in on-demand skills.