All writings

The Future of Online Education

December 15, 2022 (2y ago)

generated code

There is endless opportunity in autodidacticism.

The internet has brought about a new age of knowledge; it’s the Library of Alexandria a hundredfold, right at your fingertips - the greatest opportunity for upward mobility in the history of our civilization.

Yet online education is fundamentally broken. In this post, I’ll discuss its potential, its problems, and why the future of education is generative, personalized, and gamified.

Learning on the internet is limitless

I am a massive advocate for the movement for “open educational resources.”

It has personally advantaged me - like many others, I utilized free YouTube tutorials to begin programming in high school. I’ve taken classes on popular massive open online course (MOOC) platforms such as Udemy and Coursera, as well as from MIT, Yale, and Carnegie Mellon publicly available through platforms such as MIT OpenCourseWare.

I’ve also contributed back to this movement, creating a course directory called College Compendium in 2021 to aggregate these resources - syllabi, textbooks, lecture videos, notes, slides, and assignments - through scraping every university class a friend and I could find. The site was eventually used by many students to access courses on programming languages, machine learning, systems, security, and more.

Making knowledge more accessible and equitable; this is what the promise of education technology (edtech) is - utilizing the power of the entirety of collective human scholarship to help educate the entire world, freed from socioeconomic status, religion, or nationality. That’s what it could be.

However, it’s not what online education currently is. As The New York Times writes, “the average student in a MOOC is not a Turkish villager with no other access to higher education but a young white American man with a bachelor’s degree and a full-time job.”

I personally never fully completed a single one of those MOOCs. Of the segment of people using College Compendium we surveyed, fewer than 1% reported finishing a course.

A TechCrunch article claims that retention is “frankly, a dumb metric to condemn MOOCs over” – likening it to “the bounce rate for websites.” I couldn’t disagree more.

While retention is certainly not the only statistic to focus on, it’s emblematic of a larger problem. That problem is both simple and extraordinarily difficult to solve: learning online is (very, very) hard.

Learning on the internet is hard

The course directory I worked on was a microcosm of everything the edtech space is currently experiencing; the rapid embrace of online learning by traditional universities driven by a fear of being left behind, a desire to make education more accessible, and a failure to understand that humans are bad at learning, and the web is bad at teaching.

It’s characterized by the same challenges faced by the entire edtech space - abysmal retention, driven by a lack of motivation and academic support.

Have you ever bookmarked a tutorial or paper or course with the intent to come back later only to never return?

The majority of education nonprofits, such as Khan Academy, focus on the creation of new, high quality content. While fantastic resources for the world, I’d argue the largest challenge online education faces is not a lack of materials; there are millions of amazing educational resources already in existence.

It’s also not the dissemination of those materials; while search engines such as Google have flaws, they are servicable at surfacing information.

No, the challenge is getting people to take advantage of these resources. How is it that each of us today could spend a lifetime learning about anything to an expert level and yet we spend so little of our collective time doing so? Is it absence of interest, or awareness? Lack of understanding, or time?

Regardless, it’s not difficult to appreciate why those who are driven to begin self-guided learning fail to complete their online courses. Without the carrot and stick that is an accredited degree propped up by grades, the stakes of a large financial investment, and the engagement of a physical, full-time educational setting, many people simply lack the intrinsic motivation to keep going.

When that’s combined with a dearth of academic support - the office hours, teacher’s assistants, and tutors that help students enrolled in conventional universities become unstuck, it’s almost surprising anyone finishes.

Google is great at surfacing hyperlinks, but it’s no substitute for a teacher - there’s a fundamental difference between a link and an explanation. There’s also a lack of alignment between the business models of MOOCs and their mission; when students pay entirely upfront, there’s little incentive for the educator to care about what happens afterwards. It may be cynical, but it’s possible that this is another contributor to low course completion rates.

It doesn’t have to be this way. The future of edtech is one in which instead of trying to replicate existing educational structures online, we create new ones, designed with the strengths and weaknesses of web-based education in mind.

Learning on the internet can be personalized

Our methods for traditional education have changed very little over the last century.

Sure, now people stare at their laptops shopping online or playing Tetris instead of paying attention in class, but (especially in K-12), we’re still putting people into a classroom for several hours a day and lecturing at them until their eyes glaze over. We may be using iPads to take notes and Canvas LMS to submit work, but those are shifts of convenience, not substance.

Our system was designed for a world in which students were being prepared for a life of working in factories and on farms, in urban ovens without air conditioning (summer break!). A world in which when you needed knowledge you physically went to a library (for books, not just to study!). Clearly, the internet would fundamentally change things, right?!

So far, not really. Online courses simply replicate existing classes in a web-based format. While there are benefits to auditing such as studying anywhere at your own pace, it remains an inferior version of attending in person.

There’s an intimacy, a personal connection that exists in in-person education. Instead of failing to reproduce those qualities, we should lean into the benefits that the web provides for learning; the ability to streamline and personalize learning to an individual’s preferences.

The idea that there is a significant difference in learning outcomes for people who prefer certain styles of learning has been debunked. Yet while there’s no such thing as a “visual learner,” people can prefer different approaches to acquiring information.

I could enjoy working on a project more than listening to lectures, and vice versa. Some people are also more interested in topics than others, and so while a standardized education may be necessary for a functioning society, it’s possible to teach essential skills through a customized lens.

One student can practice writing about medical breakthroughs, while another breaks down the symbolism in “The Catcher in the Rye.” Everything else in our lives - our news, our entertainment - is now individualized (for better and worse), and our education can be too.

Learning on the internet can be generative

The advent of large language models has reshaped the possibilities for personalization. We’re using artificial intelligence to generate advertising copy, news snippets, and graphic designs. Why not apply this technology to education?

Imagine a world in which a student is learning about psychology:

Instead of the existing setup, whereby a student spends three hours watching a video lecture at 2X speed, skims some textbook pages and then copies over answers from the accompanying Quizlet into a ten question quiz, what if an AI was trained on the collective human scholarship of psychology, including every existing psychology curriculum?

That AI could create a collection of short videos for which questions are asked every few minutes; a personalized curriculum generated on the spot, tailored to a student’s interests and any academic disabilities.

This is the kind of customization currently only available to wealthy students whose parents can hire personal tutors. In less than a decade, it’s possible that everyone, everywhere, could receive this experience.

Learning on the internet can be gamified

All of this, however exciting, falls flat if you cannot motivate people to learn.

Generally, the people who sign up for MOOCs or language learning services want to master that content. Yet without the societal accountability and extrinsic motivation found in physical academic institutions, keeping that drive is remarkably challenging.

I recently read “For the Win” by Dan Hunter, which discusses the value of gamification in education and how it can alter the way we teach. Gamifying education offers immense potential in how it can replicate extrinsic motivations digitally.

Your circle of peers is recreated in an online guild, holding you accountable for your progress. Lengthy lessons are broken into smaller, bite-size sessions. Leaderboards, achievements, and certifications replace grades as symbols of your accomplishment.

Existing edtech companies have already proven this strategy works. The language-learning app Duolingo has among the best retention rates in the industry. Successful digital academic support services such as Paper, EdStem, and Piazza offer a weak substitute for office hours. Some physical schools in New York are even trying out more game-like instructional approaches.

By applying gamification strategies to more aspects of online learning, we may see increases in continued progress and overall understanding.

Learning on the internet will (eventually) change the world

I’m not advocating for improving online education with the goal of replacing universities. The university isn’t dead yet, and is unlikely to be anytime soon.

Yet while I love the experience of being a college student, I’m also conscious of the fact that there are millions of people worldwide who could benefit from an accessible alternative. In this, online education has failed to reach its potential.

The reality is, online learning and learning at a physical institution are more similar than colleges would like you to think. Universities also expect you to be effective self-driven learners; even at an “elite” institution such as Yale, the majority of what you learn comes from teaching yourself. Lectures and discussion sections are opportunities for clarification, not the source of enlightenment.

People often regard MOOCs as magic pills they can ingest, after which all of a sudden they’ll be “educated” such as in Isaac Asimov’s “Profession.” Learning doesn’t work like that. It is this where colleges excel; providing critical-thinking and the ability to grasp unforeseen concepts.

The bricks are the same both physically and digitally. If we can replicate the space between - the support structures that are the mortar connecting those bricks - we can make it easier for students to learn than it is to cheat, and render online learning more exciting and fulfilling for everyone.

We have achieved free and open access to quality education. Through making it more personalized, generative, and gamified, we can also help people actually learn it.

Thanks for reading :)