Opinion, Berkeley Blogs

The Life, Death, and Legacy of One Laptop per Child

By Morgan Ames

This is an excerpt from an interview conducted by USC Professor Henry Jenkins about my new book, The Charisma Machine: The Life, Death, and Legacy of One Laptop per Child.

In what sense was the One Laptop Per Child  “a Charisma Machine”? What are the implications of applying a term like Charisma, which has historically been so closely associated with the qualities of human leaders, to talk about technologies? Book Cover: The Charisma Machine: The Life, Death, and Legacy of One Laptop per Child by Morgan G. Ames

When I first started following the One Laptop per Child project way back in 2008, I was fascinated by how alluring the project's "XO" laptop seemed to be for many contributors and others across the tech industry. OLPC had very ambitious ideas for how its laptop should be used by children across the Global South, and what the results would be — and I found that the laptop itself came to stand for those ideas for many people. I started thinking about how the laptop began to have its own kind of authority in these circles: even mentioning it came to stand in for a particular kind of joyful, technically deep experience they wanted more children to have with computers.

I turned to sociological theory to help make sense of this, going all the way back to one of the founders of modern sociology — Max Weber — who outlined and described different kinds of authority. Charismatic authority is something that religious or cult leaders may have — they may not have the weight of an institution like the government behind them, or the weight of tradition to lean on, yet they still seem to command a following.

On the one hand, some of OLPC's leaders were certainly charismatic — Nicholas Negroponte in particular has been the public face of the project, and his charisma was important for promoting it, just as his charisma helped build the MIT Media Lab in its first two decades. But in many of the places OLPC was taken up, Negroponte wasn't necessarily well-known or, in some cases, really known at all. In these cases, OLPC's "XO" laptop itself came to stand for OLPC's ideas.

When I think about how "charisma" might apply to machines, I think about how science and technology studies (or STS) has shown that machines can have agency: they can take on meanings and act on the world beyond the intentions of their designers. I also think about how STS, and the social sciences more broadly, discuss authority not as some kind of divine or "natural" thing, but something that is produced by a whole set of social choices and technical constraints that already exist. So when I call OLPC's laptop "charismatic," it's not in a hero-worship kind of sense — it's a first step in calling attention to the ways that many have taken its allure for granted, and how that allure was created.

The MIT Media Lab has long been celebrated for its roles in “inventing the future,” yet your analysis focuses a lot on the nostalgic dimensions of the devices it created. In what senses was OLPC nostalgic? What was it nostalgic for? How do we reconcile the competing pulls towards futurism and nostalgia?

This was one of the great ironies of this project, and of many charismatic technology projects, especially in education. These charismatic projects may paint visions of a utopian future, but in order to be charismatic they have to appeal to parts of the world that are familiar to those they want to reach.

For OLPC, that was the childhood experiences with computers that many techies, especially those who consider themselves part of the "hacker" community, fondly recount from their own childhoods. In the early years of OLPC, I read through dozens, even hundreds of discussions about OLPC among project contributors and across the web that directly compared OLPC's XO laptop to Commodores, Amigas, Apple IIs, and other early computing systems that many of them had used decades before.

The specifications of these older systems were even used, in part, to justify making the XO laptop really underpowered. Reducing the laptop's energy usage was a driving goal, but the justification I heard was that these old systems didn't need fancy graphics or lots of memory to be captivating, so why does the XO need them? This ended up creating huge problems in use, though — most kids today don't really care about those older systems, after all. They want a computer that could take advantage of the media-rich web, and the XO just couldn't deliver there.

In this way, as I argue throughout the book, charisma is ultimately "conservative" — it may promise to quickly and painlessly transform our lives for the better, but it is appealing because it just amplifies existing values and ideologies. In OLPC's case, it promoted a vision of the world where children across the Global South would have the opportunity to have the same kinds of formative experiences with a computer that these adults remembered having.

You root the OLPC project in a particular conception of the relationship between technology and childhood in the thinking of Seymour Papert. What do you see as some of the core assumptions shaping this vision of ‘the technically precocious boy”?

Nicholas Negroponte was certainly the public face of One Laptop per Child, but he readily admitted in his marathon of talks in the early days of OLPC that the very idea for the project was actually Papert's, even though Papert was already retired when OLPC was announced. He often said that the whole project was "the life's work of Seymour Papert."

Seymour Papert

And when you read through all of Papert's public writing, from the late 1960s through the early 2000s, you can clearly see that connection. Papert started writing about the liberatory potential of giving kids free access to computers not long after after he joined MIT in the 1960s. Throughout the 1970s, he was a central figure in developing the LOGO programming environment. The branch he worked on, which ended up being the dominant branch, was built around the ideals of what he called "constructionism," as a tool for kids to use to explore mathematical and technical concepts in a grounded, playful way. He kept advocating these same views throughout the 1980s and 1990s, even as LOGO lost steam after many of the really grand utopian promises attached to it failed to materialize.

I argue that one of the reasons for this failure is that LOGO and many constructionist projects are built around a number of assumptions about childhood and technology that just aren't true for all children — and in fact are only true for a particular set of children, mostly boys, who have a lot of support to explore technical systems.

Some of this support comes from their immediate environment: they have parents who bought them a computer, who helped them figure it out, who were there to troubleshoot, who supported their technical interests. If it wasn't a parent, it was someone else they could turn to with questions. The programmers I've interviewed who proudly say they are self-taught had a whole constellation of resources like this to help them along.

Book cover: Designing the Creative Child by Amy Ogata But some of this support also comes from the cultural messages that we hear, and often propagate, about children. Messages about boys' supposedly "natural" interest in tinkering with machines goes back at least 100 years — there's this great volume called The Boy Mechanic: 700 Things for Boys to Do that was published in 1913! Then there's transistor radio culture, engineering competitions, and a whole host of technical toys specifically marketed to boys in the decades following. Amy Ogata, Susan Douglas, Ruth Oldenziel, and many other fantastic historical scholars have traced these histories in depth. With the rise of computing, this same boy-centered engineering culture gets connected to programming, displacing all of the women who had been doing that work as low-paid clerical workers around and after World War II, as Nathan Ensmenger and Mar Hicks have shown. The same boy-centered culture also defined the video game industry in the 1980s.

From all of this, at every turn boys — and particularly white middle-class boys — are told that they belong in this culture, that they are (or can be) naturals at programming. Everyone else has to account for themselves in these worlds, and everyone else faces ostracism, harassment, and worse if they dare to stick around. It's something I became pretty familiar with myself throughout my computer science major.

When I talk about the "technically precocious boy," it's both of these pieces — the specific material and social support certain kids get, but also the larger cultural messages they live with and have to make sense of in their own lives. This is what social scientists call a "social imaginary," or a coherent and shared vision that helps define a group.

Unless projects very actively reject and counter these social imaginaries, they ride the wave of them. One Laptop per Child is one of these, just as Papert's other projects were. Even though these projects tended to speak inclusively about "girls and boys" and "many ways of knowing," they then turned around and extolled the virtues of video games and talked about technical tinkering in ways that wholly relied on this century of cultural messaging, which had long been incredibly exclusionary.

Did this conception constitute a blind spot when applied, unproblematically, to childhoods lived in other parts of the world? How might we characterize the childhoods of the people who were encountering these devices in Latin America?

The biggest issue with relying on the social imaginary of the technically precocious boy is that the kids who identified with it have always made up a very small part of the population. If you think back to the youths of many of those who contributed to OLPC, who were discussing its similarities with the Commodores or Apple IIs of their childhoods — most of their peers couldn't care less about computers. So to assume that somehow all or most kids across the Global South, or anywhere in the world, would care when this kind of passion is idiosyncratic even in places that have long had decent access to computers is a bit baffling to me.

When I've said as much to friends who worked on OLPC, I often heard something along the lines of, "well, those past machines maybe only appealed to some kids, but this one will have much more universal appeal!" And Papert wrote about the universal potential of computers too — he called them the "Proteus of machines," with something to appeal to everyone. I see similar stories in movements to teach all kids to code.

But the majority of the kids I got to know in Paraguay — as well as those I met in Uruguay and Peru — just weren't very interested in these under-powered laptops. I found that over half of kids in Paraguay would rather play with friends or spend time with their families, and didn't find anything all that compelling about the device. The one third of students who did use their laptops much at all liked to connect to the Internet, play little games, watch videos, listen to music — pretty similar to what many kids I know in the U.S. like to do with computers. This is not to erase the cultural differences that were there, much less the legacy of imperialism still very much present across the region. But it really drives home just how wrong the assumption was that kids in the Global South would be drawn to these machines in a way that differed fundamentally from most kids in the Global North, that they'd really want to learn to program.

You note that many of the accounts that link the OLPC with cultural imperialism discount the cultural agency of the child users. What do these accounts miss?

Other analyses of OLPC provide some really insightful critiques about the project's potential effects, but nearly all of them just don't have any on-the-ground experience with how OLPC's XO laptops were actually used. Without seeing what kids were doing with the laptops, I feel like they're missing half the story — in particular, almost none of the kids cared at all about OLPC's vision or the constructionist software on the machine. If they used the laptop at all, they pursued their own interests with it.

There were also some features of the laptop — its mesh network, for instance — that generated a lot of excitement early on, but really did not work in practice. The XO laptops were just too slow and their batteries drained too fast to make mesh networking possible; the capability was in fact removed from an update to the Sugar software.

At the same time, I don't want to let OLPC off the hook when it comes to cultural imperialism. OLPC's leaders said all sorts of things about kids leapfrogging past all the adults in their lives, teaching themselves English and programming, and ultimately transforming their cultures. But what does it mean to frame kids as the primary agents of change? Negroponte often said just this in his talks: "We have to leverage kids as the agents of change." Not schools, not governments, not infrastructures, not even parents — kids.

One problem with this is that it assumes that kids are not really fully within their cultures. This fits with many of our social imaginaries of children as closer to nature and more noble for it — they aren't mired in the petty concerns of adulthood yet. OLPC hoped to capture children's "natural" interests and steer them toward computing cultures and away from the cultures they were born into. In addition to being a sneakier form of cultural imperialism, this of course didn't work — my results corroborated what researchers in education and social science know very well: that learning is socially-motivated and culturally-embedded.

A second, and much more fundamental, problem comes from the model of cultural change this promotes, which centers on children. It means that these projects are under-resourced from the beginning, because they weren't really thinking about infrastructural or institutional change — they were focused on individual change and just hoping that larger changes flow from that. And when that change fails to happen, it becomes the fault of those individuals. Failure becomes the fault of the children.


For more, you can order The Charisma Machine: The Life, Death, and Legacy of One Laptop per Child or read the full interview about the book, of which this is an excerpt: see Part I , Part II , and Part III .