Meta: early indicator of Zuckerberg’s impact on education

Last week the Chan Zuckerberg Initiative (CZI) bought a company called Meta. This small acquisition is potentially seismic in its implications for the worlds of science and education.

CZI is the philanthropic vehicle for Priscilla Chan and her husband Mark Zuckerberg (I put the couple in that order because they do). After the Initiative was announced with some fanfare, there have only been a few announcements – ambitious, striking, but mostly short on detail. Chan Zuckerberg Science was announced with the intention of “helping cure, prevent or manage all diseases in our children’s lifetime.” The highly regarded Jim Shelton was appointed head of Chan Zuckerberg Education. They announced their first lead investment in edtech and employability company Andela, alongside a group of other well-known investors in the space. So far, so expected.

The Meta acquisition is different. In the company’s own words, Meta is “a tool that helps researchers understand what is happening globally in science and shows them where science is headed.” Facilitated by the move towards Open Access in scientific publishing, the venture-backed company was selling Artificial Intelligence-driven insights about the state and potential trajectory of research to publishers, pharma companies, and others.

Now, Meta’s toolset is going to be totally free, and the company has a clear mission to open up its technology to humans and machines, and collaborate with the research community to make itself better. The for-profit drive of the company has been up-ended, towards one of impact only. The venture investors are going to take their financial exit and leave (presumably having made profits they feel happy with).

This is unprecedented as far as I’m aware. I’ve never known a philanthropic organisation acquire a venture-backed startup and mould it to its own ends. Invest, make grants, lobby – all familiar stuff from the activities of the Bill and Melinda Gates Foundation and others. Acquire – no.

The effects of this are, in my view, many. There are relatively parochial issues about what happens to the world of Scientific, Technical and Medical publishing, which I spent several years analysing with my team when at Holtzbrinck. Market leader Elsevier’s apparent strategy of building competitive advantage through amassing proprietary data and the resulting algorithms looks challenged. Clarivate Analytics (previously part of Thomson Reuters) will be worried about the long term viability of its Web of Science product.

More interesting are the broader consequences and issues. I should say first that I feel that CZI is doing something really interesting and worthwhile here, and genuinely breaking new ground. But there are wrinkles.

  1. CZI just changed a commercial market. The value chain and power relationships amongst players in scientific publishing are altered substantially – not least as this is probably just the beginning of what CZI might do, given their financial firepower. Investors and shareholders will be worried and may hold off making spending decisions. Who knows what will be changed next?
  2. There is a risk that CZI actually reduce choice and innovation in the available algorithms and tools which analyse the corpus of scientific knowledge. It’s very difficult to compete with free. I suspect they have thought of this and may well want to put money into a range of approaches, but it’s worth pointing up.
  3. Venture capitalists may now have a new exit route to add to trade sale (i.e. sale to another company) or IPO: sell to CZI if your idea is clever enough and aligns with their theses. I wonder how soon I will see this in a fundraising document.
  4. There’s a clear issue of democratic accountability here. Without malice or deliberate intention on anyone’s part, we have ended up in a situation where two people can spend huge amounts of money according to their definition of “good”, and consequently affect a major component of our society. I suspect CZI are thoughtful and well aware of this, and consulted with the scientific community before making their acquisition. Nevertheless, this is another slightly scary illustration of how polarised our world is becoming.

So what about education? All of these issues are directly transferable. I would now be very surprised if Jim Shelton and his team don’t make some major moves which re-balance power in education ecosystems, particularly in the USA but likely more broadly. Investors and existing players need to take note of whatever signals we get from San Francisco.

Yet in my view CZI need to ensure they don’t transfer their model wholesale. Science has a clear, globally accepted value system based on “standing on the shoulders of giants” – rigorous process, peer review, publication. Teaching and learning are different. As I (and many others) have written before, establishing “what works” in education is a highly subjective, value-laden exercise. CZI will need to show how they are fostering innovation and impact in education with a diverse range of cultures and contexts – or unapologetically and publicly adopt a clear set of values, and argue passionately for them. They need to do these whilst avoiding being cast as latter-day cultural imperialists and facing a backlash. This is a tall order in a “post-truth” world where Zuckerberg has already struggled with issues of editorial responsibility in his day job at Facebook. The team have awe-inspiring ambition, power, and potential, which they have only just started to deploy. Things just got interesting.

An open operating system for education, the threat of Big Tech, and how progress might happen

In my twenty years of working in education technology around the world, I have become convinced of the compelling case for an “open operating system for education”. If digital tools of all sorts were able seamlessly to exchange data without human intervention, but according to clearly defined rules governing privacy and permissions, the potential would be hugely exciting. Public, not-for-profit and commercial players in the system could compete and collaborate on the basis of how well they improve teaching and learning, not their access to proprietary data. There would be room for many more experiments and innovations from large and small organisations, as an open source algorithm written in a back room could be compared to $100m interventions from mega-corporations. We would have a much better chance of knowing what worked, why, and in what context. Governments, administrators, teachers, learners and their parents could better control and analyse the data they create. The ecosystem could be fairer, more transparent and vibrant.

We’re certainly not at that point today. Such a system would rely on a set of comprehensive, widely adopted, open standards for how digital educational tools should interoperate. Vitally, these standards would have to be agnostic to commercial model or local cultural norms (for example, around privacy). They would necessarily support as many choices as can be thought of. Not because all choices are good choices in my or your opinion, but because decisions in the value-laden world of education are usually context-specific, culturally debatable, and often develop over time. Texas would make different choices to Turin or Tianjin, and these all need to be upheld.

There are a lot of standards out in the world, but they aren’t complete and they often focus on very specific American contexts. This isn’t the place to examine the alphabet soup of LTI, LRMI, SCORM, QTI, Caliper, OneRoster, ed-Fi and more, but suffice to say not all of the people developing all the standards talk to each other or have the resources to consider the world outside the USA. Additionally, the standards aren’t universally supported by tools and are sometimes resisted by vested interests. Indeed companies such as Clever are working to help sort out the spaghetti of proprietary systems by offering a (proprietary) interoperability layer. Plenty of data is being created by some fascinating organisations– from AltSchool to JISC to Zaya Learning – but their full potential is untapped.

For me, one issue now dominates whether or not we will make progress towards this open operating system. It is the role of the huge technology companies which are increasingly reaching into every part of our lives – Amazon, Apple, Google/Alphabet, Facebook, Microsoft, Alibaba, Tencent and Baidu.

The tech titans are on a mission in education. Google Classroom. Recently, Microsoft Classroom too. And Apple’s additions to iOS 9.3 so that it works better in schools. And Amazon’s recent release of Inspire. And Alibaba’s investment in TutorGroup. And a raft of other developments over the last few years as education technology has once again become a focus for investors, corporations and policymakers across the world.

Regrettably, these huge companies don’t seem to have considered some of the long-term consequences of their moves (or possibly don’t care about them). There is a clear and present risk that their actions will stifle innovation, create unintended controversy, and (most importantly) reduce the impact that we can all make as stakeholders in the world of learning.

Some background. Taking just the US players, Alphabet/Google, Amazon, Apple, and Microsoft are each pursuing a “Trojan Horse” approach in the small-for-them global education market: they’re loss-leading in line with their wider strategies.

  • Google continues to improve the functionality of its free Classroom tool for schools, universities, teachers and students. Classroom is part of the Apps for Education suite and extends it so that users can not only send and receive email and work on documents together, but assign work to classes, collect students’ grades, and discuss homework. (In edtech jargon, it’s a free LMS). Chromebooks – which run a Google operating system – are now the bestselling devices to US schools, likely as they are cheap and easy to manage. They require a Google account. All of this pulls in data and users: Google’s raison d’être.
  • For many years, Microsoft has made Office free for schools and universities. They hope that this will ensure that people use their productivity tools out of habit and familiarity for the rest of their lives. Newly-launched Microsoft Classroom – very similar to Google’s offering, if a few years behind the curve – is part of free Office 365 for Education.
  • Apple is fundamentally a hardware-driven business. It likes to sell iThings and Macs and for people to become used to them (and so keep buying more hardware, apps and content). So, selling beautiful devices into education is good for Apple on all fronts – and if this requires some tweaks to iOS, so be it. Apple are even prepared quietly to purchase an edtech startup to further their strategy in education.
  • Amazon wants to be the distributor of as much as possible. Digital content especially – films, TV programmes, ebooks, audiobooks. Education resources have thus far been out of their reach as publishers and others have their own direct-to-schools sales staff. So they have bought a company which creates mathematics content, created a site for teachers to discover and share resources, and offered “one stop shop” solutions to cities and (allegedly) entire nations.

All, you may say, fair enough: isn’t it brilliant that teachers and learners across the globe get free or reduced price tools from these big players which let them do their jobs more efficiently and cheaply? And surely part of the companies’ motivations is about helping out the world of education?

Yes and Yes. But.

But – data. Google, Microsoft, Apple and Amazon lock up the data created in their own proprietary systems. They have absolute control over who gets access to it, for what purpose. Some access to the rich seams of information created by teachers, learners, administrators, content, applications (and more) is possible, but right now none of these solutions abide by any widely adopted open standards for data exchange. Furthermore, it’s rarely entirely clear who owns the data and what it can be used for, by whom.

Right now, this isn’t too much of a problem. Big tech’s offerings are relatively immature and little used in comparison to their competitors such as Blackboard, Desire2Learn’s Brightspace, Instructure Canvas, Moodle, Edmodo and many more. The competing offerings support some standards with varying degrees of usability, and in many cases do the best they can.

However, the scope and functionality of the Google, Apple and Microsoft solutions are inexorably growing. These tools are free at the point of use and are backed by relatively trusted brands with immense marketing clout. This means that such software is easily – and sometimes unthinkingly – adopted in the educational context, where budgets never go far enough. There are cases of entire countries’ education systems deploying Office 365 for Education. Furthermore, it seems that many adopters are sleepwalking right now – at a recent educational conference I asked three informed and intelligent teachers on a panel whether or not they were concerned that Google was in control of their gradebook data. Their responses implied – “Google Classroom is free – what’s not to like?”

In short, as adoption of these gratis tools unavoidably accelerates, educational data and access to it may become increasingly controlled by a small powerful group of private companies based in the USA. Google Classroom et. al. are of particular concern not because I think the big technology companies will necessarily do anything bad – I’m no conspiracy theorist. But we need to ensure that at least the majority of data created by education is public and personal rather than commercial and proprietary, and the track record so far of the big tech companies in opening up data is – shall we say – mixed.

Public, open standards are gnarly, painful to create and maintain, glamourless, and invisible if they work well (nobody thinks about TCP/IP as they “like” something on Facebook). But they are sometimes essential for the health and development of the system which they underpin. I for one firmly believe that educational data is one of those instances.

So what to do? It would be nice to think that Amazon, Microsoft and their peers could see the public benefit of collaboration. Perhaps uniquely, they have the capabilities to knit together the current patchwork quilt of standards into a workable whole and seamlessly support them (they have just about agreed on HTML). It’s debateable whether or not open standards would harm their business models – whilst “lock in” would clearly be reduced, users might be more prepared to use all of their offerings sequentially or even simultaneously – a teacher using Google Classroom on an Apple Mac to access a quiz made in PowerPoint bought through Amazon Inspire. Arguably, such openness would accelerate the adoption of their free tools even more.

Sadly, there is currently little evidence that these issues are even on the agenda of the technology giants. As with other markets such as music, at some point we may find ourselves presented with a fait accompli which seems beyond the control of governments. Control may already have slipped away.

In my view we need to work to avoid this. In the absence of Big Tech stepping up to the plate, plenty of bottom-up and top-down initiatives could help to ensure educational data is properly available and managed. Governments and groups of schools and universities should only accept contracts where there are clear assertions of data ownership (ideally by the learners themselves) and access (for both learners and institutions). All market players could accelerate the existing work which is going on by actively engaging with standards initiatives and encouraging them to work together. As teachers, learners, parents and citizens we all need to hold vendors and institutions to account. If we don’t, education will likely be the poorer.

For references and links, contact me in the comments or on Twitter @nkind88