I’ve been writing about the potential of data and technology to transform society for 15 years. Re-reading my earlier material is, frankly, pretty cringe-worthy. “Sunlight is the best disinfectant,” we digital optimists said in 2008, clutching copies of Clay Shirky’s Here Comes Everybody while imploring people to open up their datasets and let the “wisdom of the crowd” do its work. The crowd wasn’t looking so wise a few months later during a global financial crisis that few predicted. This week, as Russia used another disinformation campaign to justify invading Ukraine, one recalls the hope we once pinned on social media during the Arab Spring.

Looking back, I have not been not alone in my misplaced optimism. In 2004, Bill Gates predicted email spam would be a thing of the past, within two years. In 2014, the EU Commission swallowed Facebook’s (now Meta’s) ludicrous assertion that matching Facebook IDs and Whatsapp users’ mobile phone numbers was a technological feat impossible to man. The $122m euro fine the Commission eventually leveled closed the stable door on a proverbial horse that had bolted.

The international development community is not immune. From the Taliban seizing US biometric devices to UNHCR’s Rohingya data crisis, to the World Food Programme’s Palantir woes and the World Bank’s Doing Business Report’s downfall, the development data community has sometimes hoped for the best without planning for the worst. These failures are compounded by a tendency to concentrate decision-making in the global North while overlooking the concerns of people who have the most to lose when things go wrong.

We must not abandon hope. To course-correct we should stop grasping at each new advancement as a blank slate, telling ourselves that this time will be different. No data innovation or technological advancement exists in a vacuum. To ignore the biases that shape them and their applications is to set ourselves up for failure. Asymmetric power begins with the narrow life experience of the people who develop the devices, apps, and algorithms that increasingly govern our lives. “It has become appallingly obvious that our technology has exceeded our humanity,” Einstein said—half a century before machine learning made that statement doubly true.

The diverse contributions to the Data Values Project are a sign that the data revolution of the last decade is evolving into something more intentional and determined. Hundreds of organizations in our network are engaged in constructing new data ecosystems not beset by the issues of the broken systems we all encounter daily. To avoid pitfalls, we can start by doing three things:

  • Make agency and dignity the cornerstones of development data projects. The ideals of the Data Values Project are global, but the best solutions will always be local.
  • Call out bad faith actors — including the subset of CEOs who cloak their mercenary tendencies in idealistic language.
  • Challenge government technocrats to let communities shape their data collection processes with consultation, participation, and representation of the people whose lives will be affected by data collection efforts.

This is easier said than done. There are structural challenges like time poverty—where the voices we must hear are squeezed out by their lack of time to participate in civic life. Disenfranchisement and a lack of funding also hamper good faith efforts at data inclusion. These power dynamics are complex and only partially understood. That’s why we need a movement that starts by stating, “it doesn’t have to be this way,” and empowers us all to talk honestly about our failings. It’s also why we must collaborate for positive change. Only then will our cautious optimism be grounded in reality while inspired by our utopian origins.