Caribou Digital turned five in 2019. When we started back in 2014, the world felt very different. Election results were different, technological challenges were different. It’s fair to say that some of that era’s enthusiasm about technology’s role in society has dissipated. That said, the urgency surrounding the challenges of digital inclusion and the transition to digital economies has never been more real.

Over those five years we’ve been documenting our evolving thinking, and the result is a list of five things we’ve learned, and five things we think will matter for the next five years.

Most digital development projects still exist in sectoral silos, or are driven by a relentlessly optimistic approach to “connecting the last billion.” This is despite all the evidence pointing to the significant impacts of digital technologies on economies and societies. So why is the development industry so slow to change?

Development tends to love binary problems. You’ve either had a polio vaccine or you haven’t. You’re either earning above the poverty line or not. We can agree on evaluation processes that provide detailed data to prove the success of our programs. We can use randomized control trials to prove the effectiveness of our solutions. Once we’ve created a binary problem we can quantify the challenges we face in terms of large numbers that play well as shareable social media assets. The last billion, etc.

The flaw with viewing digital connectivity as a binary problem is that, although the system is built on binary logic, tangible benefits the system provides to real people are not. The questions shouldn’t end with “How many people in the world are connected to the Internet?” We need to ask: “To what kind of internet are people around the world connected?” Or: “How effectively are they using it?”

For these reasons, development organizations and foundations wishing to engage with digitization have to avoid simple binaries (such as advocating for closing the digital divide). Instead, they need to engage with complex, nuanced concepts around agency, literacy, power, and structure in the digital age.

Connectivity is a prerequisite for inclusion in the global information economy and for the functioning of 21st-century states and institutions. But connectivity is not enough in and of itself. We must champion it, but we can’t expect better cell connections or social media to kickstart development on their own. These are tools, the underpinning structure on which prosperity can be created, but they are not themselves drivers of prosperity.

At Caribou, we’re tired of simple equations that show “If you connect X number of people to technology Y you achieve Z% increase in GDP.” We’re bored of reading business models that use US-centric assumptions about the attention economy to predict that we can end poverty by giving everyone control over their data. We are equally sceptical of simplistic discussions around the potential income benefits of digital connectivity and digital work.

That’s not to say we should throw these arguments out completely. Digital work will undoubtedly provide new incomes and increase the quality of livelihoods for many, but, even in the US, JP Morgan has shown that the average Uber income dropped 53% between 2013 – 2017 as the platform matured. How will gig-work platforms grow and change in other markets? There is clearly value in the data economy, but our own research on digital advertising in emerging markets shows how little poor people’s digital data is worth— Facebook’s most recent billion subscribers in low-income markets only earn around 50-80 cents per quarter gross income per user.

Digital has the power to unleash huge untapped reserves of human capital. Enabling this will require investment in digital infrastructure, but digital platforms—the glue that connects digital infrastructure and human capital—are sucking value out of both states and citizens.

From digital financial services that encourage saving more than gambling, to social media platforms that encourage self-expression and connection instead of misinformation and acrimony, our clients and discussion partners are increasingly interested in examining and promoting digital inclusion that includes ethics, dignity, and self-actualization.

Skills gaps and new definitions of digital literacy are coming into sharper focus as we discuss the future of work in the digital age. Digital literacy can’t be achieved by teaching a few more folks to code. In a world where literally everybody on earth needs to have a sense of how to interact with each other via platforms, we need to promote social and platform digital literacy, so that people know how to conduct their digital lives, from education and expression, to livelihoods and citizenship, safely.

Even as the cost of internet access drops, the ongoing project to create a world that supports inclusive and meaningful use will require many approaches. Of course, there’s an interplay between startups, technological innovations, capital and the private sector. But the broader challenges are interdisciplinary, including psychology, sociology, political science, human computer interaction, economics, and ethics.

Power matters. Decisions made in Palo Alto or Seattle or Beijing alter user experience for millions of people. Markets are becoming platforms. Yet the onus for change (and response) can’t lie entirely with individuals or their skills. What role should philanthropy, aid, and regulation play in shaping the structure of the internet and the markets and social sectors that it shapes?

There is exciting and energising work happening in this area. (For example, work by CIS on zero rating and net neutrality in India.) These conversations at the intersection of technology and governance are examples of critical pieces of the puzzle.

But sensible digital policy is notable by its absence. For better or worse, the traditional role of western policymakers in providing global leadership is struggling to keep up with the scale and rate of technological change. Partly because of the changing nature of global politics, and partly because the time gap between digital technologies being adopted at scale in the Global North and the Global South has collapsed. The old “direction of travel” for policy leadership has gone. Outside of GDPR, the US and EU are doing no better at regulating global digital platforms than anyone else.

When other countries see empty chairs in Congress and EU hearings where platform senior management should be sitting, it suggests we’re all on our own. Worryingly, parochial, or even authoritarian, regulation is a viable option in this vacuum of leadership. So while Uganda implemented social media and mobile money taxes, to detrimental effect, many countries have simply banned access to the internet and services during elections or crises. We fear some countries may find the restrictive example of China, which forces concessions from global platforms, increasingly attractive.

Speaking at an event at the excellent Andela Nairobi Campus in 2019, we told the assembled programmers how lucky they were to see the world in a new way. We chose our words carefully—they see the world differently through phenomenal amounts of available data. We didn’t talk about data as the new oil or a new asset class, but as something that programmers and data scientists see. Or more bluntly, what they choose to see and what they don’t.

Smartphones, Satellites, and Sensors are opening up the way we see the world in whole new ways. On the one hand, they provide an enormous opportunity to establish more inclusive, holistic, and global approaches to understanding how humans connect to nature and to each other in socioeconomic systems.

On the other hand, the flourishing conversations about the ethics of AI, bias in algorithmic computation in the Global North, must be brought into the development community. It’s critical that we focus on not just the data we can gather, or even on its potential, but rather on the choices we all make to deploy, process, interpret, and govern it. Each new network innovation — telephony, television, roads, internet, etc. reminds us of the interdependence between a network’s potential and its political economy: who is incentivized, who holds power, and who benefits.

A full treatment of the interdependency between private control and public good is well beyond the scope of this post, but it’s worth flagging that as sensors, imaging, and artificial intelligence augment the already amazing potential of the internet we have today, how will these new capabilities best benefit the public, not just in the Global North but around the world?

As a tangible example of this compression and globalization of the challenge of shaping new technologies for good, consider ‘digital identity’. It’s become clear to us that the challenges states face in granting digital identities to the one billion people who lack a them are intertwined with the challenges of preventing surveillance, tracking, and erosion of privacy. There are a billion people who are seen too little by the emerging digital infrastructure, and at least a billion who now feel they are seen too much. Good policymaking and good innovation will keep the tension between these issues in mind. One can’t be solved without the other.