Kentaro Toyama’s critique is a welcome turn in the public discourse on the potential for technology to address major challenges of our time, such as economic development and democratic governance. In contrast to alternatively Elysian or Orwellian prophecies about the effects of ICT, Toyama’s account is sober and nuanced.

Seeing firsthand the high ratio of failures to successes in developing countries, Toyama transformed from technological optimist to skeptic. Perhaps so many unused and rusting computers should not have been surprising, though. Even in the high-motivation, high-capacity environment of Silicon Valley, many more projects fail than succeed. Given lower capacity, it would be miraculous if the batting average in India were better.

But what matters is hits and runs rather than the batting average. As long as someone invents an iPhone for every few Apple Newtons or Microsoft KINs, technology will continue to improve human welfare.

Toyama’s concern, then, is whether the batting average can be increased. He urges that we focus less upon technology itself, and more on those meant to use it. In particular, he argues that technology only amplifies people’s underlying motives and capabilities. This is surely correct, except that motivation and capacity cannot be understood apart from particular technologies. The fact that my three-year-old son has no difficulty navigating to his favorite game on my iPhone and playing it illustrates this intimate connection.

Intentions often become clear only after some technology is deployed. Those who brought mobile phones to Nigeria and other African countries probably did not guess that one of the major benefits would be their ability to disseminate market information. And those who brought telecenters to India did not guess that they would be used primarily for gaming and pornography (though developed countries might have provided a clue).

In other words, users might not know what technology is good for until they use it. As Henry Ford reportedly quipped, “If I had asked my customers what they wanted, they would have said, ‘a faster horse.’”

While motivation and capacity are no doubt necessary for technologies to deliver on their promise, efficacy—whether a technology helps to accomplish some developmental objective—is a third and perhaps even more crucial consideration. What are the problems that telecenters or plastic laptops are supposed to solve, and how are they supposed to solve them?

The first question of efficacy is whether some piece of ICT can help at all. ICT can provide new information, accelerate data manipulation, and facilitate communication. These contributions are enormously useful in a large number of scenarios, but not all. In a low-tech environment, ICT will not help plough the fields or secure potable water. It may be that, as a general matter, even developmentally appropriate ICT is less useful in societies where human well-being depends more upon moving and making things than upon manipulating data.

In recent years, tech-savvy activists have helped the disadvantaged gain political voice.

A second question of efficacy is what besides ICT must be in place in order to solve some social problem. Even when technology figures in the solution to political or economic development, it enters only as one piece of a larger puzzle in which the other pieces are often missing. Mobile phones help farmers, for example, when market information is the missing piece and others—water, fertile soil, safely traveled roads connecting farms to markets—are in place.

In the provocative second part of his essay, Toyama applies concerns about the digital divide to development. Far from especially benefitting the poor, ICT, he argues, generally exacerbates social and economic inequalities in developing countries. He reasons that those who are well off usually have greater capacity to utilize technologies and will reap greater benefits from it. And this critique only applies to general-purpose technologies. Some technologies serve more specialized—usually wealthy—audiences and needs. As Toyama points out, many engineers and the companies that employ them will design tools to serve customers who can pay.

However, recognizing that technology can be biased by design to benefit some socioeconomic groups over others leads to a constructive possibility. Redoubled efforts to create technology that will be of special value to the least advantaged might help to offset technology’s tendency to heighten inequality. Think of this approach as affirmative-action engineering.

An affirmative-action effort of this sort might help the disadvantaged gain political voice. In recent years, tech-savvy activists have done just that, developing tools to enhance democratic integrity and political accountability. Ushahidi—a crowd-sourced platform first deployed to monitor and map incidents of election violence in Kenya—is the most prominent of these tools. Ushahidi has been used in diverse crisis situations such as violence in Gaza and the earthquake in Haiti. To the extent that such technologies increase democratic control and responsible politics, they benefit those who are less well off.

Another kind of affirmative-action effort would enhance the quality and accessibility of public goods. The poor often rely to a greater extent on public goods such as state-funded primary education, governmentally constructed infrastructure, and public security. Projects such as Kiirti in India and Cidade Democrática in Brazil create ICT platforms that facilitate reporting on problems with public goods. They also make it easier for reformers to agitate on behalf of improvements.

So any particular technology is only one element of a potential solution to some problem of economic or political development. We should make sure other elements are in place before investing great faith in ICT. And the proper response to Toyama’s caution regarding the inequity of technology is to expand our efforts to develop applications for the disadvantaged. As Toyama says, intention matters. And it matters not just for those who use technology, but for those who create it as well.