Smarter, crowdsourced, super startup schools

Ben Williamson

Disruptive innovation. Bleeding edge. Scalable solutions. The Uber for X. Silicon Valley is routinely ridiculed for the language of technology entrepreneurship and startup culture it has dispersed. Yet the Silicon Valley lexicon is fast becoming part of the language of education, and is galvanizing significant financial and R&D practices around educational technology. Tech companies are investing with unprecedented optimism in new educational technologies, with an estimated $1.6billion of venture capital invested in educational technology in the first half of 2015 alone, following a massive upward trend in ed-tech funding over the previous 5 years. But they are also increasingly using their massive financial power to create their own new schools. In the last few years, IBM has launched P-TECH, a network of ‘smarter schools’ that are modelled to fit its Smarter Cities global program. A former Google executive has established AltSchool, a chain of schools designed more like makerspaces than conventional schools. The founder of Khan Academy has opened his own ‘Lab School’ as an experimental site. And the widow of Steve Jobs of Apple has dedicated a huge philanthropic donation to a school redesign competition, XQ: Super School Project. Rather than tinkering in the margins of state schooling to increase efficiencies and effectiveness, Silicon Valley is seeking to disrupt the established model of the school, to embark on a kind of creative destruction of the institution of education itself.

These innovations are, to borrow a phrase, the ‘Uber for School.’ They are proposed as radically disruptive innovations in education. I call them ‘startup schools’—new kinds of educational institutions that originate in the culture, discourse and ideals of Silicon Valley startup culture, and that are designed to relocate its technical practices and politics to the whole social, technical, political and economic infrastructure of schooling. These new schools are being designed as scalable technical platforms, underpinned by software engineering expertise; they are funded by commercial and venture capital and philanthrocapital sources; staffed and managed by execs and engineers from some of Silicon Valley’s most successful startups and web companies; and proposed to reinvent, reimagine and rebuild education in the mould of Silicon Valley itself. And though they remain primarily US-based, the conditions in other countries may be amenable to their expansion—a new global market in Silicon Valley school solutions. Startup schools are designed to run on lines of code; they also reflect the social, cultural, political and economic codes that regulate the ‘culture of code’ of tech entrepreneurship.

Smarter schools
P-TECH (Pathways in Technology Early College High School) is a chain school model for high school education, designed and promoted by the IBM Corporation. Its emphasis is on vocational education for careers in science, technology, engineering and maths (STEM). Many students are offered internships at IBM itself during the course of their studies, and to date some of its graduates have continued to full employment with the company.

The model originated in 2011 at P-TECH Brooklyn in New York, through a partnership of the New York City Department of Education, The City University of New York, New York City College of Technology and IBM Corporation. It was the result of meetings between New York’s former Chancellor of Education Joel Klein and IBM’s then-CEO Samuel Palmisano in 2010, and conceived as a way of connecting education and employment, in particular to equip students with the same skills and degrees that IBM demands of its own employees. By 2015 it had expanded to around 40 schools in the US, with plans for expansion into Australia after a visit from its Prime Minister. Barack Obama himself has visited a P-TECH school and declared it a successful model. The mayor of Chicago has ‘ordered’ 4 P-TECH schools for the city after meetings with the New York state governor. There is strong political support, as well as tech expertise, for the P-TECH model. It has even been profiled in Wired magazine, which noted the careers pipeline into the tech sector represented by its approach:

Tech companies are long on excuses about why they’ve been so slow to diversify their ranks, even in the face of constant criticism. But by far the most frequently cited reason is they can’t hire diverse employees en masse until the country builds a diverse pipeline of skilled tech workers. With P-TECH, IBM has done nothing if not create a prototype of that pipeline. Now, it’s calling on other tech leaders to take that prototype and do what they do best: scale it to the millions of people—in this case kids—who need it most.

The article also notes that P-TECH fosters an explicitly competitive, entrepreneurial ethos, one well suited to the tech sector itself.

Notably, IBM has produced extensive documentation on the P-TECH model, a kind of rule book on how such a school should be established, funded, organized and run. But it does not necessarily run all 40 schools itself: Microsoft and other tech companies, plus engineering, manufacturing and healthcare systems organizations and groups have all taken responsibility for different schools in the chain. Local city and state budgets are responsible for the day-to-day financing of the schools, but IBM has itself invested significantly in the development of curriculum materials, technical infrastructures, advocacy and promotion, and public visibility for P-TECH.

Less visibly, P-TECH is also an outgrowth of IBM’s much wider ‘Smarter Cities’ global program, and in particular the Smarter Education strand of it. The IBM Smarter Education program is based on claims about the real-time availability of educational data and its usefulness in school improvement:

Schools and universities have always recorded and stored data as they tracked grades, attendance, test scores and demographics. With the increasing availability of technology in the instructional process, educational institutions now collect, in real time, data about what their students learn and how they progress … using big data and analytics.

Like the smart city itself, its ‘smarter schools’ are ‘enhanced through data, mobile and cloud technology,’ including a mix of ‘analytics, mobile, social and security solutions built on cloud infrastructure to monitor academic progress of individual students.’

In one of the few critical pieces of coverage of P-TECH, Greg Linday claims it is intended ‘to build for schools what its operations center is for cities: a single system for collecting, aggregating and analyzing data from students and teachers alike, then writing algorithms to prescribe how to cope.’ He claims that P-TECH is mobilizing a ‘software “infrastructure layer” for schools, running behind the scenes to manage students’ digital textbooks and analyze their performance,’ and that P-TECH ‘is a research project for gleaning best practices that can be codified into software or peddled by IBM’s consultants to other clients — in this case, schools.’ P-TECH schools ultimately act as laboratory sites for IBM to test out its analytics capacities, and to realize its imaginary of Smarter Education, as well as a research site for the production and piloting of software products that might be commercialized. P-TECH schools are built in IBM’s data-driven image.

Makerschools
A slightly different model to P-TECH is that of the ‘makerschool.’ The idea of the makerschool has emerged in part from the growing homeschooling culture in California. A recent article entitled ‘Hacking Education’ in Wired has detailed that many Silicon Valley coders, hackers and makers are now choosing to educate their own children through the DIY logic of digital making. It profiles a new breed of homeschoolers—the techie parents who see public or state education as fundamentally broken, and have chosen instead to educate their children themselves. The DIY homeschooler sees makerspaces and hackerspaces as ideal kinds of educational institutions, where children are learning directly through tinkering, hacking, coding and making, rather than being educated in the prescriptive, standardized mould of schools–a kind of digital age mashup of progressivism usually associated with John Dewey and the ‘unschooling’ of John Holt.

The new makerspace-schoolers acknowledge, however, that such an individualized model can’t work ‘at scale.’ The intriguing technological solution offered by the unschoolers is to create new kinds of ‘hybrid’ schools, somewhere between homeschooling and traditional school.

Perhaps unsurprisingly, the solution has been found by a new crop of startups using technology to create new educational models. The prominent example is AltSchool, set up in 2013 by Max Ventilla, a former tech entrepreneur and Google exec, which ‘prepares students for the future through personalized learning experiences within micro-school communities.’ Its stated aim is to ‘help reinvent education from the ground up.’ After establishing in four sites in San Francisco as a ‘collaborative community of micro-schools,’ AltSchool expanded in September 2015 to Brooklyn and Palo Alto, with further plans for new schools in 2016. It has since hired executives from Google and Uber (both of which are also involved in education through the HackingEdu annual hackathon that ‘helps developers revolutionize the education industry’) plus other successful Silicon Valley startups. The AltSchool chief technology officer, formerly the engineer in charge of the Google.com homepage and search results experience, has stated that ‘I am highly motivated to use my decade of Google experience to enable the AltSchool platform to grow and scale.’ Elsewhere on the AltSchool site, the AltSchool ‘platform’ is described as a new ‘central operating system for education,’ a scalable technical infrastructure that can be transported to new sites.

AltSchool brings the startup model to school. Its website refers to ‘technology-enabled models’ that are transforming industries and institutions, such as Airbnb, and applies these ideals to education. As a tech platform, AltSchool is managed on analytical, technical and scientific lines, albeit laced with the progressivist discourse from which it draws its central philosophy. As the AltSchool values claim:

Our personalized learning approach puts each child at the center of everything we do … coupled with state-of-the-art classroom design and technology, [and] a flexible learning environment that mixes individual, group and experiential learning.

Our analytical approach and core strengths in innovation combine educational best practices with the latest tools. Our educators build learning experiences that are adaptive at their core and keep our children engaged.

We can anticipate the model of makerspaces and hackerspaces migrating to other kinds of makerschools, startup schools and pop-up schools that, run on experiential learning principles, encourage greater exploration, inquiry and problem-solving through the active construction of knowledge and understanding, whilst monitoring and regulating the experience through learning analytics and adaptive learning software. School is broken, they seem to be saying; let’s hack it, and create startups to do so. As one interviewee in the ‘Hacking Education’ piece in Wired states, ‘The cost of starting a company has gone down because there are online tools you can use for free. I can see that happening with school. So much of that stuff is just up for grabs.’

Setting up a new community of micro-schools, however, is not inexpensive. AltSchool originally raised $33million in venture capital funding, with a further $100million investment in 2015, including donations from Mark Zuckerberg. AltSchool is, then, thoroughly governed, managed, and financed through the discourses and material practices of Silicon Valley startup culture. Its technical infrastructure as a platform is modelled on social media. Its funding is almost exclusively generated through venture capital and tech philanthropy. Its engineering and design team are applying their social media expertise in data dashboards, algorithmic playlisting, adaptive recommender systems, and app development to the development of new ed-tech devices and platforms. If you look closely at the photos on the website, you can read a sign on the wall of AltSchool HQ that reads ‘Be Altsome.’

Experimental R&D lab schools
Khan Academy provides thousands of hours of online tutorials and videos to millions online, offering ‘practice exercises, instructional videos, and a personalized learning dashboard that empower learners to study at their own pace in and outside of the classroom.’ Not content with remodelling education as online personalized learning, its founder, Salman Khan, launched Khan Lab School in September 2014. Located in Mountain View, in the SanFrancisco Bay Area, near Google HQ, Lab School is intended to realize the vision of schooling Kahn had previously outlined in his 2013 book The One World Schoolhouse. In the book Kahn makes an epochal claim for a historical shift in the organization of schooling:

New educational institutions and models emerge at inflection points in history. Harvard and Yale were founded shortly after the colonization of North America. MIT, Stanford, and the state university systems were products of the Industrial Revolution and American territorial expansion. We are now still in the early stage of an inflection point that I believe is the most consequential in history: the Information Revolution. And in this revolution, the pace of change is so swift that deep creativity and analytical thinking are no longer optional; they are not luxuries but survival skills. … I composed a mission statement that was both wildly ambitious and—with the help of readily available but absurdly underutilized technology—completely attainable: Provide a free, world-class education for anyone, anywhere.

As an institutional and pedagogical instantiation of this mission, Kahn’s Lab School teaches math, literacy and computer programming–in line with its tech sector roots–but also emphasizes ‘real world’ projects, personalized learning, student-centred learning, and a strong commitment to building children’s ‘character’ and ‘wellness’ through, for example, mindfulness training.

Most notably, however, Lab School has been established as an experimental R&D lab for testing out different educational approaches and technologies, and aspires to contribute to the production of new theories of learning itself. Lab School has been profiled in Wired, which noted that its

goal isn’t just to build one fancy school but to develop and test a new model of learning that can be exported to other schools around the country and the world. His team is diligently recording and tracking every student’s progress and sharing the findings with their parents and the staff, an open source approach to educational innovation. In this view, the Lab School kids are guinea pigs … willingly subjecting themselves to new ideas that have never been tried before, then adapting and adjusting and trying again.

“This is a lab for establishing new theories that could affect the
rest of the planet,” Khan says. “The whole point is to catalyze change.”

Like AltSchool above, Lab School’s ‘touchy-feely surface’ of character education ‘masks a rigorous fealty to tracking data about every dimension of a student’s scholastic and social progress.’ In particular, it uses data analytics to provide a constant and growing trace of the character development of its pupils, and reinforces those data through standardized testing.

It also welcomes outside organizations in to the school to test out new ideas and technologies, so that the children are positioned as constant willing subjects of a tech experimentalist approach. AltSchool likewise engages its students in regular HCI-based experiments to develop, test and finetune its operating system.

The educational operating system of these startup schools is a mishmash of data analytics and character education. They assume that character can be broken down into measurable indicators. The eduOS of such schools depends on weaving a whole bunch of character theory from psychology with data analytics. This vision is well illustrated by the AltSchool ‘recommended reading list’, which features a list of texts emphasizing the importance of educational concepts such as ‘whole-child learning,’ ‘personalization,’ children’s ‘character,’ ‘growth mindsets,’ ‘self-control,’ ‘resilience,’ and ‘creating young innovators.’ These texts are all drawn from recent thinkers on character education, psychology, educational neuroscience, and digital learning—texts which provide a conceptual vocabulary for learning.

The image of school children as innovative, self-controlling characters with growth mindsets posited by such texts is woven into both Kahn Lab School and the AltSchool platform, which is described on its website as ‘the magic that allows teachers to architect a personalized educational experience based on each student’s unique circumstances and learning objectives.’ What this means is that the data platform acts to perform a continual audit of ‘whole-child learning’ by enacting analytics processes that can mine pupil data for indicators of their character, mindsets, resilience and so on, and then automatically customize or personalize curriculum, pedagogy and assessment around their individual needs. AltSchool and Khan Lab School project a front door that emphasizes new wave psychological concepts of growth mindsets and character development, plus new age ideas about wellness and mindfulness (alongside ideas with longer historical lineages from progressivism and unschooling), which they then track and monitor through surveillant back-end analytics platforms to build data profiles of their students’ character. The pedagogies they design in response to these data-based character insights can then be seen as experimental methods for subsequently intervening in the development of children’s character–a kind of R&D of the person.

Crowdsourced super schools
And now we have XQ: Super School Project. Set up in September 2015 with $50million philanthropic investment by Laurene Powell Jobs, the wife of Steve Jobs from Apple who died in 2011, the XQ Super School Project is conceived as a massive ‘democratic and crowdsourced’ experiment in school design. It is conceived as an open competition ‘to reimagine and design the next American high school’ in order to ‘deeply prepare our students for the rigorous challenges of college, jobs and life.’ Launched with a significant social media campaign including high-production video content and a celebrity launch event modelled on a new product launch for an Apple device, Super Smart School is based on the claim that schooling has remained static for a century as a ‘dangerously broken system,’ while the wider society has experienced waves of technological innovation. The project is soliciting proposals in late 2015, with the objective of partnering with winning teams to provide them with expert support, including the allocation of $50million funding for the winning 5 proposals to turn them into ‘real Super Schools.’ In addition to the main competition, the project is also crowdsourcing ideas for school innovation through the Twitter hashtag #RethinkHighSchool and is running a roadshow in autumn 2015 to encourage design teams to gather together, discuss their ideas and get input from the project team.

Despite being an open, crowdsourcing competition, Super School Project is designed with a number of clear constraints. For a start, it assumes that neuroscience is the best place to start in understanding learning processes. A paper on the ‘science of learning’ provided on the website refers to ‘understanding and applying the fundamentals of brain science’ to ‘empower young people to become agents of their own learning journeys.’ It draws on neuroscientific claims about the malleability and plasticity of the adolescent brain, about the brain-based nature of students’ ‘mindsets,’ and particularly applies these to the idea that Super Schools should aspire to ‘foster a mathematical mindset.’ In another paper on the skills students require in the 21st century, the Super School Project dismisses the so-called ‘old paradigm’ (of following orders, being product-driven, 9-5 lifelong employment and domain specialization), and replaces it with the ‘knowledge economy’ paradigm of co-creation, distributed leadership, flexibility, domain agility and creativity. These 21st century skills have become a kind of mantra in the tech sector. Brain science and the knowledge economy are two of the dominant discourses informing the Super School Project, and ultimately act as design constraints for potential entrants to the competition.

The Super School Project is managed by the XQ Institute, itself an incubated product of the Emerson Collective, a philanthropic organization that claims to ‘invest in ideas and fuel innovation’ through partnering with entrepreneurs. Its founder and president is Laurene Powell Jobs (one of the richest women in Silicon Valley, and the world’s 9th wealthiest woman) and its managing director is Russlynn Ali (a former assistant secretary in the US Education Department). There is serious financial Silicon Valley power twinned to political power in education policymaking embedded in the Super School Project.

Code/school
Rob Kitchin and Martin Dodge have influentially introduced the concept of ‘code/space’ to describe spatial environments that depend on computer code for their functioning. The startup schools being developed by Silicon Valley entrepreneurs such as IBM’s P-TECH model, Salman Khan’s Lab School, and the AltSchool platform, are ultimately new kinds of educational code/spaces running on a technical infrastructure—instantiations of an emerging form of ‘code/school.’ Code/school articulates the kind of educational institutions that thoroughly depend on coded infrastructures and devices to function as intended and designed. This could be a fragile arrangement. The history of tech innovation is itself littered with failures. Software problems, miscodings and ‘hasty hacks’ could give rise to ‘buggy, brittle’ services, with potentially catastrophic consequences, as Anthony Townsend has argued. What might one little miscoding cost when startup schools start scaling up?

But Kitchin and Dodge also assert that understanding how code functions means looking beyond the lines of code themselves. It means examining the programmers who script code to accomplish specific goals; the professional and institutional cultures of coding in which code is written; the financial and business arrangements that fund it; the social and technical infrastructures in which it is embedded; the management and governance structures that regulate it; and, not least, it requires close attention to the work that the code does in the spaces in which it performs. There is a massive and urgent research agenda around the code/spaces of startup schools that could be addressed through such lines of inquiry.

Moreover, it is the language of Silicon Valley which is now increasingly infusing education, and supporting the kinds of coding practices that will make startup schools operational. The Silicon Valley discourse of innovation, entrepreneurship, startup culture, makerspaces, crowdsourced solutions, platforms and philanthrocapital is becoming a new language of schooling. State schooling is, of course, not unproblematic. But it is being problematized by the tech entrepreneurial sector according to a very particular style of thinking, with its own discourse, explanations, arguments and solutions, as well as its own objectives. For IBM, it is clear that P-TECH is intended at least in part as a talent pipeline for its own purposes, but it also sits very neatly with its wider smarter cities solutions that it is selling commercially to urban governments around the world. The P-TECH curriculum is the perfect selection for educating the smart city. AltSchool is expanding its own ‘operating system’ to new sites, implanting the language and material practices of the tech sector into schooling.

Startup schools are also completely changing the governance structures of schooling to include tech entrepreneurs and software engineers in the management teams of schools. Schools are increasingly being governed by tech expertise, financed by web giants, and organized through the software products they produce. These actors not only have their own technical ways of doing things; they also see the world in particular ways, and identify problems in ways that can be solved through particular means, or ‘silicon bullets,’ to borrow a term from Tom Liam Lynch. The politics of educational governance are being completely transformed. Local and national governments no longer have responsibility for schools; education governance is happening in Silicon Valley HQs, operationalized by software engineers, and financed by corporate, venture and philanthrocapital funding. Language, or discourse, generates practices, and their operationalization has consequences. It makes some ideas and objectives seem possible, desirable and attainable, while silencing others. Silicon Valley discourse is becoming the operative syntax for the future governance of schooling.

Though P-TECH, AltSchool, Khan Lab School and XQ Super School Project are all US initiatives, the conditions are right for them to extend elsewhere. P-TECH is already slated to move into Australia. And the UK government ‘free schools’ program might accommodate a ‘micro-school’ model like that of AltSchool. In terms of governance, UK school boards of governors are already being transformed into more businesslike, entrepreneurial outfits, and there is more commercial involvement in the curriculum through the academies program. The opening of startup schools outside of the US would allow Silicon Valley to govern education at a distance, projecting its values and ambitions through its globally distributed networks of HQs and infrastructures into schools from afar.

As a final point, as startup code/schools are owned by big tech companies, any student data generated by these institutions also belongs to those institutions to use to conduct various analytics procedures. This circumvents wider concerns among parents and critics about access to student data by third party commercial organizations. Instead, startup schools have direct access to their data, and can collect and calculate it in-house–a significant example of the ‘capture model‘ of data collection that allows computers to track information in real time, identify particular human activities, and reorganize the data sets in ways that can be used for intervention. In this way startup schools act much like the social media companies from which they are derived, whose business plans depend on the capture and analysis of customer and user data, often with little external scrutiny, for the purposes of better profiling and prediction of individuals’ habits and social trends. At a time when emotional manipulation of users by Facebook and search engine manipulation of public knowledge by Google are becoming major concerns, the implantation of data analytics in the everyday functioning of code/schools—tracking and predicting everything from attainment to behaviour and emotions in classrooms—should itself be the topic of close scrutiny. How might startup schools collect data, who will own it, what will they do with it?

So while startup schools may be conceived as ‘angel investments’ from Silicon Valley to the social institution of education, they also need to be understood as part of a complex culture that is rooted in the logics of technical innovation, commercial business planning, and social media data capture. It is not just the language of Silicon Valley that is penetrating education through startup schools; these schools represent new forms of patentable innovation, intellectual property, and commercial investment for the tech sector. Smarter, crowdsourced, super startup schools are also surveillant, data-capturing, experimental laboratories and scalable venture capitalist schools built to run on the social, cultural, economic and political operating systems of Silicon Valley itself.

[Note: I updated this post with new material on Lab School on 28 October 2015]