From Books to Bytes
A Learning Revolution for the Poor
Many friends, colleagues and family members were puzzled by my decision to leave a successful career in finance to dedicate myself to Rumie back in 2013. It’s hard to communicate the magnitude of the learning revolution that is underway today, and the opportunity we see to empower those who have traditionally been denied the chance to realize their innate potential. So I’ve tried herein to place it in the broader context of human history — beginning with a look at a transformative innovation from over five hundred years ago.
The Limits of Literacy
At the beginning of the 15th century, spreading knowledge was difficult. Texts needed to be laboriously copied by hand. They sometimes contained errors — which were then recopied in future versions. And since books wore out easily and were vulnerable to the elements, an army of scribes was required to maintain a sizeable collection of books.
Johannes Gutenberg’s invention of the printing press during that century began a revolution in the dissemination of information. Using typesetting technologies from Asia, his “movable type machine” made the mass production of texts possible at a reasonable cost for the first time. As the cost of books dropped and their availability increased, literacy levels climbed upwards and the Renaissance began soon after — as humanity’s most precious resource, brainpower, started to be more thoroughly exploited.
Despite the mass production of printed books in subsequent centuries, the cost of accessing written knowledge still exceeds the grasp of many of the world’s poorest today. Stories abound of African classrooms in which one or two outdated and dilapidated textbooks are shared amongst an entire classroom of children. While printed materials are cheaper, better, and faster than pre-15th century written materials produced by scribes, they still entail material costs for pulp and paper, transport costs to get them to remote areas, hefty intellectual property costs such as royalties to authors, and profit margins for retailers and distributors.
As a result, humanity’s greatest resource (again: brainpower) remains untapped in many parts of the world. By some estimates, there are 1.8 billion school-aged children around the world, of which more than half, one billion, lack anything resembling a quality basic education. These brains are like plants which, if fertilized and watered, could grow into a generation of minds that could rapidly advance progress for some of humanity’s most pressing problems. Instead, this human potential remains undeveloped.
Digital Revolution
In the latter half of the 20th century, computers were invented and became increasingly widespread in society. First we had mainframes. Then later came minicomputers. Not long after that, workstations and PCs spread into mainstream society. Next came portable laptop computers. And now we are in the age of ultraportable mobile smartphones — which, despite being small and relatively cheap, each hold greater computing power than all of NASA had back in the 1960s.
Each new generation has heralded a change in scale: mainframes were just for companies. Later on, PCs entered the home and were shared by the family. Eventually they became cheap enough that families could afford more than one. Cheap laptops have enhanced this. And now smartphones have yet again changed the game, pushing the boundaries as their low-cost, affordability, power-efficiency, and mobility have made them truly personal devices in a way no computing device has ever been before.
At the same time, data transmission costs have dropped rapidly over time as improved compression techniques and massive investments in telecommunications have reduced the cost to deliver data. Whereas the marginal cost of pulp and paper has changed little in the last 20 years (indeed it has increased slightly), the marginal cost to deliver a byte (a unit of data) has plummeted over the same period. Once the infrastructure of a device and an internet connection is in place, a written text can be transmitted at a de minimis cost even to remote locations.
The rise of the Internet has thus facilitated easy sharing of information for lots of people around the world. One result of this trend is that resources to build skills — ranging from how to write software code to basic business concepts — have become easier to access, as those with knowledge and subject area expertise share their insights digitally and freely under open licenses.
While some of this free content is of questionable quality, much of it is of even better than the expensive alternatives of the past. One study showed that students in Utah using free, open source textbooks from one non-profit provider scored statistically higher than similar students using traditional $60 K12 textbooks. More and more examples are appearing of freely available materials being proven to be of equal or greater quality than the expensive materials of yesteryear.
And as early internet pages have moved from text and basic images to now real-time high definition video streams, the dissemination of learning information online has gone far beyond replacing written materials and now encroaches on classroom lectures, supplemental out-of-classroom learning, interactive and adaptive learning that emulates tutors and teachers, and even one-on-one human interactions that take place remotely via voice and video conferencing solutions. The need to socially distance in response to the spread of COVID-19 globally has accelerated some of these trends, but they were slowly building year over year well before the pandemic.
Thus far, much of this has been taken for granted: giving well off people things they already have isn’t all that exciting. (After all, most rich societies already had access to that information in some form: perhaps at a well-stocked public library, or through textbooks and lectures from free publicly-funded schools and after school programs, or even on bookshelves at home.) And while it certainly is more convenient to have learning content in digital form — interactive, easier to distribute and update, and less weighty in a child’s backpack — it’s not all that transformative to move from well-funded, good quality learning via paper to well-funded, good quality learning via pixels on a screen.
Even having free access to an encyclopedia, which used to cost $1,000 to purchase, is barely noticed. Though folks appreciate the convenience of being able to look up anything from anywhere (it’s easier to settle a dinner table bet on the main export of Botswana if someone can quickly pull out a phone and check in seconds), few appreciate today that the drop in effective cost that has come alongside this digital shift now makes it possible to deliver these superior resources to the majority of people worldwide, ranging from youth in Detroit to children in Afghanistan.
Those still left behind
Yet the digital revolution has not yet reached most of the world’s poorest. Many have pondered the question: how can educational technology be used to rapidly improve the lives of the world’s poorest? Many a TED Talk has pontificated grandly about the potential for this, alongside other grand ideas, but few have articulated a useful method by which this would come about. While the idea is nice, so far it has remained far too expensive and too difficult to implement in practice.
One initiative bravely attempted to build a special device to bring technology to the world’s poor children. One Laptop per Child (OLPC) announced in 2005 that it would bridge the digital divide for the world’s poorest by designing and manufacturing a low-cost, durable laptop that children from Ghana to Guatemala could use — equalizing the playing field.
Unfortunately, OLPC didn’t really work out. It took longer to design and build than promised. It cost far more than expected — over $200 per device, a price point too expensive for many target markets. The hardware didn’t work particularly well and was soon eclipsed by better, commercial alternatives. And most problematic of all, there was no content strategy, since most of the leading sources of free learning content didn’t exist when large OLPC projects launched in 2007 and 2008 — nor did its designers have much interest in content, focused as they were on the idea of delivering a specially designed computer-for-the-poor to a child as an end in itself.
Ultimately, OLPC was widely considered a disappointment. Third party studies on large scale, expensive deployments in Peru found that students using OLPC devices showed no measurable gains in math or reading scores compared to students who did not have them. It turned out that giving them a specially designed device with no content and a poor internet connection had little impact on their learning outcomes.
Timing is Everything
OLPC was a noble effort, but it was an idea before its time. In technology, timing is everything. Research firm Gartner has advanced the idea that new technologies often go through what it calls the “Emerging Tech Hype Cycle,” wherein a peak of inflated expectations amongst media and investors gives way to a trough of disillusionment — markets overshooting, as they are wont to do — before finally the initial promise is realized.
This same phenomenon occurred in the early 2000s, as basic mobile phones started to spread into developed countries. Many dreamers recognized the promise of bringing such devices into emerging markets, where the lack of landline infrastructure meant that a ‘leapfrog’ to mobile phones would be much more transformational to societies and to economic growth. Unfortunately, many companies bought wireless spectrum and built networks when equipment costs were high and the cheapest handsets were well over $200. Many companies thus failed or were later forced to restructure debts. They failed to realize that the technological capability alone wasn’t enough for developing countries — the leapfrog innovation also needed to be cheap enough.
But some five or six years after developed countries reached high levels of mobile phone penetration, the scale economics behind producing global cellular networks and hundreds of millions of handsets led the costs to plummet and a $50 handset became possible. And this finally led to the beginning of the mobile revolution in emerging markets. OLPC’s foray into education came shortly after and, while noble, did not heed the most important lesson of the mobile leapfrog in developing countries: to spread in these countries, one doesn’t need flashy, cutting-edge technology; one needs good technology with cutting edge economics.
Rumie and the Digital Learning Revolution
Twelve years ago I worked on investments into mobile phone networks in emerging markets when handset costs reached $50, believing that the true promise of mobile phones was not taking the world’s richest from something good to something great, but rather taking most of humanity from nothing at all to something great.
Similarly, I believe that the major human impact of a proliferation of high-quality, adaptive digital learning tools won’t be the marginal gains for a minority of humanity who are already quite well-educated; but rather a transformative leapfrog for the vast swaths of humanity who have never had access. That conviction led me to found Rumie in 2013 with another leapfrog in mind: from traditional offline learning methods, including in-person classroom-based learning and expensive paper-based materials, to a digital approach that is cheaper, better and faster. In other words, a revolutionary transition from ‘books to bytes’ for the world’s poorest.
Although we don’t offer hardware today (Rumie works on any device), in the early days many communities lacked access to a low-cost device. So in the early days at Rumie, we began by working with an existing tablet manufacturer to modify a $50 tablet, jerry-rigging it with extra battery life and extra storage space so that it could store an entire learning curriculum comprised of high-quality open source materials, allowing it to work even without internet access. Internet access would allow syncing and updating materials, but the goal was that when fast internet wasn’t available — meaning the vast majority of the time — the device would not become a paperweight. (Simply giving up on trying to create a fast, live internet connection where it didn’t yet exist aggrieved many technology utopians, but with our practical focus we didn’t want to be Steve Jobs; for our target markets — with all the infrastructure and resource constraints they faced — we figured the late 80s TV character Macgyver was a better inspiration.)
We started expanding through partnerships with local partners in various countries using primarily English language materials and saw excellent results: it turns out digital learning materials that show statistical gains for kids at an already high baseline in rich countries can be even more impactful when absorbed by kids who start with test scores some 40–50% lower but are just as capable. Based on feedback from partners on the ground, new features were added: user profiles (in recognition that no matter how on-the-ground programs were designed, the devices were inevitably being shared amongst students and family members); gamification (so kids could be rewarded with some game playing time by teachers and administrators as a reward for good behaviour); and analytics (to pave the way for the kind of real data analysis, impact testing and measurement that was never possible before in paper-only societies).
By 2015, it was clear that this technology was affordable and was driving cost-effective impact in many diverse regions and countries — ranging from villages in Guatemala to Syrian refugee camps. But while the hardware and operating system were usable in many places, the educational content was not. (While most markets want some some ‘international’ resources in English, Guatemala and Syrian refugee camps needed resources in Spanish and Arabic too.) And organizing content to meet such criteria is hard: there’s too much of it from too many online sources in English alone, to say nothing of covering diverse subjects that may be useful to these communities across so many different languages and age levels. An organization with 10 or 20 people could not possibly do it; one would need an army of 10 or 20 thousand, or likely far more.
So in 2015 we introduced the LearnCloud, an online platform on which volunteer educators come and meet requests from NGO partners to organize content from the myriad free resources available and produce curriculum in specific subject areas and languages —rapidly populating a giant and open repository of free learning content that is collaboratively organized and localized for different communities and cultures. We figured that if Wikipedia could leverage non-profit status to build such a formidable army of volunteer workers — such that it put its for-profit competitors out of business — why couldn’t we leverage the wisdom of the crowd behind the noble goal of educating the world’s least fortunate adults and children?
The results were impressive. Volunteers as far away as the Arctic joined in, using technology to contribute to the education of children in Syrian refugee camps in Turkey. (Seriously: there were passionate volunteers at Arctic College in Nunavut, way up in Northern Canada.) It turns out that folks are a bit fed up of always being asked by NGOs to open their wallets and give money, money, money to help causes far, far away. When asked, they’re more inclined to generously donate their time, skills, and passion first — if there’s a way to do that, that is.
Does History Rhyme?
Mark Twain once said that “history doesn’t repeat itself, but it does rhyme.” Around 15 years ago, after developed countries had enthusiastically adopted mobile phones, scale economics drove the cost of the cheapest handset toward $50, unlocking a similar growth pattern in wireless penetration in emerging markets. Today, roughly a decade after folks in rich countries enthusiastically adopted smartphones and tablets (which largely share the same components and software), the cheapest tablets and smartphones on the market have dropped to $50 — making them affordable for widespread use in poorer countries.
The GSM Association, the industry organization that represents the interests of mobile network operators worldwide, said in its 2020 annual report that by the end of 2019 almost half the world was using mobile internet (nearly 4 billion), a number that was increasing by 250 million people per year, and that nearly three-quarters these mobile internet users were in what are called ‘low and middle-income countries’ (or LMICs, for those who are into international development jargon). In addition, this was driven in part by the decline in smartphone costs “from 44% of monthly income in 2018 to 34% in 2019, driven primarily by increased availability of lower cost devices in Sub-Saharan Africa and South Asia.”
This proliferation of billions of smartphones globally may be a reason for concern: the environmental waste created by devices flooding the planet is a legitimate problem. That’s why we partnered with EcoATM, the largest device recycling company in North America and Western Europe, to help learners with access to cheaper, recycled devices that are a win both for them and the planet. (We even jointly hosted a pop-up shop in Manhattan in early 2020 in which people could ‘sell’ their used phones to EcoATM, which in turn donated them for use in our programs for girls’ and womens’ education in Afghanistan.)
If smartphone penetration continues to grow in these markets — ideally in an environmentally responsible way — the infrastructure that will result makes the ‘books to bytes’ transition not only cheaper, better, and faster, but frankly inevitable. But a device and an internet connection is still not quite enough; it’s necessary but not sufficient. Barriers still remain to perfecting the formula for this free digital learning revolution. One of them cited in the same GSMA report is a lack of digital literacy and skills, which is cited as the biggest obstacle by mobile users in these markets. (This now beats affordability, which came in second place.)
We knew about this obstacle many years ago: having spent the first five years of our existence building projects with local partners around the world that impacted hundreds of thousands of learners in over 30 countries, ranging from remote indigenous communities in North America to the Philippines, and all 100% of them using mobile devices (even when totally offline), our data and experience yielded a few key insights:
- People don’t learn on mobile devices in the same way that they do in the offline world; blindly ‘digitizing’ offline solutions does not work well, as many have learned the hard way when the COVID-19 pandemic forced in-person learning programs to go virtual almost overnight.
- Mobile users prefer applications that are not only mobile-optimized, but mobile-first, and strongly prefer micro-experiences (a few minutes at a time) over long, involved commitments (watching a one hour lecture).
- The greatest area of demand voiced by the communities themselves was in job, life and career skills. As a technology organization, we listen closely to feedback from our users; it also fits with our anti-paternalistic approach.
Based on these insights, in 2019 we started a strategic review to build on our first five years of unique experience delivering free mobile learning around the world, evolving our solution to meet where we think the world is going:
- Focus on practical, applicable skills: We shifted our content focus to centre primarily on life, career and job skills going forward — normally considered ‘supplemental’ or out of classroom learning — rather than on any core high school (K12) educational curriculum. Besides the fact that it was actively demanded by communities ranging from refugee camps to urban centres in rich countries, the need is clear: various studies have outlined a growing ‘skills gap’ in the workforce in many parts of the world.
- Evolution to Microlearning: People don’t use mobile devices the way they use other technology (laptops), or offline options (books). Rather
than spending hours watching lectures or reading, they prefer short snippets of information that can be consumed quickly while on the go, in those ‘boredom’ periods that social media companies expertly hoover up. (The average instagram session is just over 6 minutes, but it adds up: total social media usage in the US is around two and a half hours per day, and is often higher in poorer countries.) Besides the fact that recent studies, such as one in the Journal of Applied Psychology and another from Dresden University in Germany, argue that microlearning is roughly 20% more effective than traditional learning methods, by delivering information via snippets on mobile devices it means that people are far more engaged and use more of it — which helps us improve our quality and social impact. - Expanding through Mobile Operators: With financial backing from the GSMA, we’re currently growing a partnership we started with Roshan Telecom, the largest mobile operator in Afghanistan, to bring bytes free of data charges to any of their over six million subscribers. The project is modeled after a similar partnership they launched with Wikipedia a few years ago. Our Afghanistan projects have a focus on girls’ and womens’ education, given the cultural and safety barriers they often face there; by working with the mobile operator we can effectively bypass many of these obstacles and directly empower the learner through their low-cost mobile device, not to mention scale this to millions across the country instantly. (As an added bonus, our latest solutions work on any device, including basic feature phones too.) We’re now working on scaling this model with other mobile operators.
Volunteerism remains at the centre of the Rumie ethos. A growing volunteer community is at the heart of our work, much like Wikipedia, with two key differences: (i) our growing community of volunteers appears to be far more diverse than that of Wikipedia (well over half are female, for example), which we believe will result in a better, more representative and locally-relevant library of learning resources; and (ii) we vet all the content before it publishes online. Celebrities have joined in too, such as this byte on understanding curiosity by astronaut Chris Hadfield. And so have corporate volunteers: as described in this post on virtual volunteering, since the COVID-19 lockdowns began in March 2020, employees at Manulife, The Carlyle Group, Amazon and others have contributed through fun virtual volunteering sessions in which they distilled their insights into new bytes that teach job skills that are aligned to their areas of expertise.
As all of these myriad trends unfold at once, Rumie is poised to grow its impact exponentially. Just as dispersed volunteers online created Wikipedia, the world’s largest encyclopedia (60 times the size of Britannica and growing), Rumie’s fast-growing volunteer community of skilled volunteers and subject area experts is building the world’s largest repository of free, mobile-first microlearning content online. And like Wikipedia, we’re structured as a non-profit, meaning Rumie will always be freely accessible to everyone, everywhere.
Over the next decade we will witness the beginning of the greatest revolution in access to information since Gutenberg introduced the printing press in the 15th century, as those who didn’t have books leapfrog directly to a cheaper, better and faster method: bytes. And finally, humanity will rapidly begin to harvest vast unused tracts of its greatest and most precious renewable resource. We hope the Rumie Learn will play a valuable part in this revolution.
P.S. For more on how we’re directly targeting social media usage, see this recent article I wrote: From Social Media to Social Impact. Our goal is to help people give their brains a diet, replacing a dopamine rush that causes damage to mental health with a dopamine rush from learning a new skill.
(This article was updated in 2021; it was originally written in 2016.)
