Take a photo of a barcode or cover
mburnamfink 's review for:
The Charisma Machine is a hard-hitting deconstruction of the One Laptop Per Child project, conducted through the brutally unfair techniques of writing down what proponents of the OLPC program, primarily Nicholas Negroponte of the MIT Media Lab claimed it would do, and then looking at what children in the developing world actually did with the machines.
OLPCs at a primary school in Kigali, Rwanda in 2009, from Wikimedia
The goals of the program were quite ambitious, hundreds of millions of laptops for the world's poorest children, running open source software, made of durable and easy to replace parts. The laptops would inculcate these children into MIT's remix-reprogram-remake hacker culture, recreating the idyllic childhoods of the current technological elite, where an early childhood spent getting programs to run in BASIC turned into successful engineering careers. The OLPC project served to link a doable project of designing, building, and distributing low-cost laptops, with ongoing global concerns about education, nostalgia for the garage start-ups of the early PC days, and glossy TED talk futurism. All of this was backed up with the constructionist educational theory of Seymour Papert, where the computer was the perfect tool to learn to think with, in concert with elegant but limited programming languages like Logo. In Ames' theoretical framework (which I'll return to), this made the XO hardware a charismatic machine, capable of activating transnational networks to save the children.
Negroponte and other OLPC boosters had visions of tough laptops literally being dropped out of airplanes and used by eager children, but real world use required on-the-ground partners. Ames did her fieldwork in Caacupé, Paraguay, a district capital in a developing nation, where the NGO Paraguay Educa was heavily invested in making the OLPC project a success. Paraguay Educa installed wifi at test schools and provided teacher training and after school activities. They were thoughtful and well-resourced. They had the best of intentions.
But the friction of the real world is a far cry from the ideals of the program. The first bit of friction was that the OLPC was a profoundly limited machine, with 256 MB of RAM and 1 GB of Flash memory. Processor specs are roughly equivalent to a 2000-era PC, with the miniscule storage going back to 1995. In 2005, when the XO was proposed, the specs were bad. In the web-centric world of 2010, when Ames' primary fieldwork was conducted, the specs were crippling. Users complained that the OLPC took long minutes to book up, and the battery life was barely an hour when it was on. Even though Paraguay has good electrical infrastructure for a developing country, there were not enough outlets in homes and classrooms to enable long-term use of the OLPC by all students. Her description of a classroom exercising involving the Tux Paint app, which decays into a fiasco of tech support woes, is all too familiar. More friction was added as students removed apps to store more music or video content, and automatic OS updates deleted student projects. Rather than inspiring ownership, the broader context inspired a sense that these were someone else’s machines.
By Ames's calculation from Paraguay Educa stats, roughly 1/6th of XO's became unusably broken within a short time, with screens a particular problem. Beyond that, her ethnographic survey showed that fully one half of students in the program did not use their XO's beyond the minimal mandated activities, preferring to focus on non-digital activities like soccer, socializing, or helping around the house. Of the one third of program participants who were active users, the most common activities were playing music (Daddy Yankee was particularly popular at the time), watching cartoons, and playing emulated games via WINE, including Mario. Students became most adept not at the constructionist programming tools, but at pushing the limits of the hardware to make the OLPC a multimedia toy. My favorite part was her description of OLPC principle Walter Bender’s visit to Paraguay, and how a carefully stage-managed performance was used to cast him in the role of the technocratic patron. It was a display worthy of Secretary McNamara touring a strategic hamlet to demonstrate progress against Viet Cong infiltration.
A handful of students did become adept users, and Ames follows these exemplars in detail. But a closer looks reveals that these students had exceptionally supportive parents, and generally came from the upper reaches of Paraguayan society. Breaking into the ranks of international hackers requires English proficiency to learn a real programming language like Python, which in a country like Paraguay already means being part of a cosmopolitan elite. If the OLPCs had any effect, it was rendering the computer a little less of a novelty for participants. Perhaps now, nearly a decade on, some of Ames’ respondents are entering the workforce with a sense of how to use basic office programs, google for help on a technical problem, and generally not freeze when confronted with new technology. I’m skeptical any are programmers today. Measured against its original goals, OLPC met almost none of them.
The fieldwork is dedicated and detailed. As a researcher, Ames has a remarkable talent for objectivity. The simple facts are damning enough. If this book has a flaw, it’s in the STS theoretical paradigm, which takes up the first chapter. Ames blends Weber’s charismatic authority with Jasanoff’s sociotechnical imaginaries and Latour Actor-Network Theory to argue for the OLPC itself being charismatic. And there is a sense in which certain technologies (hyperloop, deep learning, blockchain!) warp discourse around them, when there are much less sexy realities (trains, linear regression, SQL) which actually solve those problems. This is very much an STS dissertation, which means that it has to push STS theory in a new direction. It’s just that, if I may get on a soapbox as somebody with an STS PhD for a moment, the gap between the objective of STS as a field, to think critically about the relationship between humans and technology, and the rigor and usability of major paradigms in the field is so deep that the best use of most of the theories is being buried in a hole in a desert.
The heavy STS theory is a shame, because Ames uncovered a far better theoretical paradigm in nostalgic design, in this case the tendency of the OLPC leadership to create something for their idealized childhood rather than the actual lived experience of children in the developing world. A second theoretical lens is the role of psychological theories in educational programs, and particularly the weaknesses of Papert’s constructionism for real classrooms and real students. As in most of these things, I blame Reviewer #2.
Griping about the discipline aside, Ames’ fieldwork is exceptional, the writing clear, and while the case-study is perforce limited to its specific site, the results are extensible to any number of ‘one clever design hack to fix a complex sociotechnical problem’ charades.
(Disclosure Notice: I received a free copy of the book from the author, and no other compensation.)
OLPCs at a primary school in Kigali, Rwanda in 2009, from Wikimedia
The goals of the program were quite ambitious, hundreds of millions of laptops for the world's poorest children, running open source software, made of durable and easy to replace parts. The laptops would inculcate these children into MIT's remix-reprogram-remake hacker culture, recreating the idyllic childhoods of the current technological elite, where an early childhood spent getting programs to run in BASIC turned into successful engineering careers. The OLPC project served to link a doable project of designing, building, and distributing low-cost laptops, with ongoing global concerns about education, nostalgia for the garage start-ups of the early PC days, and glossy TED talk futurism. All of this was backed up with the constructionist educational theory of Seymour Papert, where the computer was the perfect tool to learn to think with, in concert with elegant but limited programming languages like Logo. In Ames' theoretical framework (which I'll return to), this made the XO hardware a charismatic machine, capable of activating transnational networks to save the children.
Negroponte and other OLPC boosters had visions of tough laptops literally being dropped out of airplanes and used by eager children, but real world use required on-the-ground partners. Ames did her fieldwork in Caacupé, Paraguay, a district capital in a developing nation, where the NGO Paraguay Educa was heavily invested in making the OLPC project a success. Paraguay Educa installed wifi at test schools and provided teacher training and after school activities. They were thoughtful and well-resourced. They had the best of intentions.
But the friction of the real world is a far cry from the ideals of the program. The first bit of friction was that the OLPC was a profoundly limited machine, with 256 MB of RAM and 1 GB of Flash memory. Processor specs are roughly equivalent to a 2000-era PC, with the miniscule storage going back to 1995. In 2005, when the XO was proposed, the specs were bad. In the web-centric world of 2010, when Ames' primary fieldwork was conducted, the specs were crippling. Users complained that the OLPC took long minutes to book up, and the battery life was barely an hour when it was on. Even though Paraguay has good electrical infrastructure for a developing country, there were not enough outlets in homes and classrooms to enable long-term use of the OLPC by all students. Her description of a classroom exercising involving the Tux Paint app, which decays into a fiasco of tech support woes, is all too familiar. More friction was added as students removed apps to store more music or video content, and automatic OS updates deleted student projects. Rather than inspiring ownership, the broader context inspired a sense that these were someone else’s machines.
By Ames's calculation from Paraguay Educa stats, roughly 1/6th of XO's became unusably broken within a short time, with screens a particular problem. Beyond that, her ethnographic survey showed that fully one half of students in the program did not use their XO's beyond the minimal mandated activities, preferring to focus on non-digital activities like soccer, socializing, or helping around the house. Of the one third of program participants who were active users, the most common activities were playing music (Daddy Yankee was particularly popular at the time), watching cartoons, and playing emulated games via WINE, including Mario. Students became most adept not at the constructionist programming tools, but at pushing the limits of the hardware to make the OLPC a multimedia toy. My favorite part was her description of OLPC principle Walter Bender’s visit to Paraguay, and how a carefully stage-managed performance was used to cast him in the role of the technocratic patron. It was a display worthy of Secretary McNamara touring a strategic hamlet to demonstrate progress against Viet Cong infiltration.
A handful of students did become adept users, and Ames follows these exemplars in detail. But a closer looks reveals that these students had exceptionally supportive parents, and generally came from the upper reaches of Paraguayan society. Breaking into the ranks of international hackers requires English proficiency to learn a real programming language like Python, which in a country like Paraguay already means being part of a cosmopolitan elite. If the OLPCs had any effect, it was rendering the computer a little less of a novelty for participants. Perhaps now, nearly a decade on, some of Ames’ respondents are entering the workforce with a sense of how to use basic office programs, google for help on a technical problem, and generally not freeze when confronted with new technology. I’m skeptical any are programmers today. Measured against its original goals, OLPC met almost none of them.
The fieldwork is dedicated and detailed. As a researcher, Ames has a remarkable talent for objectivity. The simple facts are damning enough. If this book has a flaw, it’s in the STS theoretical paradigm, which takes up the first chapter. Ames blends Weber’s charismatic authority with Jasanoff’s sociotechnical imaginaries and Latour Actor-Network Theory to argue for the OLPC itself being charismatic. And there is a sense in which certain technologies (hyperloop, deep learning, blockchain!) warp discourse around them, when there are much less sexy realities (trains, linear regression, SQL) which actually solve those problems. This is very much an STS dissertation, which means that it has to push STS theory in a new direction. It’s just that, if I may get on a soapbox as somebody with an STS PhD for a moment, the gap between the objective of STS as a field, to think critically about the relationship between humans and technology, and the rigor and usability of major paradigms in the field is so deep that the best use of most of the theories is being buried in a hole in a desert.
The heavy STS theory is a shame, because Ames uncovered a far better theoretical paradigm in nostalgic design, in this case the tendency of the OLPC leadership to create something for their idealized childhood rather than the actual lived experience of children in the developing world. A second theoretical lens is the role of psychological theories in educational programs, and particularly the weaknesses of Papert’s constructionism for real classrooms and real students. As in most of these things, I blame Reviewer #2.
Griping about the discipline aside, Ames’ fieldwork is exceptional, the writing clear, and while the case-study is perforce limited to its specific site, the results are extensible to any number of ‘one clever design hack to fix a complex sociotechnical problem’ charades.
(Disclosure Notice: I received a free copy of the book from the author, and no other compensation.)