Self Help

Internet for the People - Ben Tarnoff

Author Photo

Matheus Puppe

· 38 min read

“If you liked the book, you can purchase it using the links in the description below. By buying through these links, you contribute to the blog without paying any extra, as we receive a small commission. This helps us bring more quality content to you!”



Here is a summary of the key points in the preface:

  • The internet has a physical infrastructure that spans the globe, like the MAREA submarine fiber optic cable that connects Spain to Virginia. This shows the internet is not just virtual - it has a material, historical body.

  • The infrastructure of the internet often follows older networks built for political and economic control, like colonial telegraph networks and Cold War telephone lines. The internet continues this political tradition.

  • The internet is commonly seen as “broken” today due to issues like misinformation, surveillance, and monopoly power. Proposed solutions tend to focus on increased regulation and breaking up big tech companies.

  • But the root issue is that the internet is owned by private companies and run for profit, so an internet run by smaller, regulated firms would still have problems. The internet was not always this way.

  • The book will examine how the internet was privatized, starting as a public research network and transitioning to a corporate-owned system optimized for profit. Reversing this process through public or cooperative ownership may be necessary to fix the internet’s problems.

  • Understanding the physical, political history of the internet shows it was constructed through human choices and can be changed through democratic struggle. Its current problems are not inevitable.

  • On November 22, 1977, a van carrying a computer drove between San Francisco and San Jose as part of an early experiment with packet switching technology that would enable the Internet.

  • The van transmitted data packets via radio signals that were amplified by mountaintop repeaters. The packets then traveled through telephone lines across the country to satellites and receiving stations that relayed them across the Atlantic Ocean.

  • After bouncing between multiple locations in Europe, the packets returned to the U.S. and made their way to a computer in Marina del Rey, CA, just 400 miles from where they originated.

  • This landmark experiment demonstrated how data could be broken into packets, transmitted through a “network of networks” along variable routes, and reassembled by the destination computer.

  • While the content of the message has been forgotten, the significance was in proving how networks could interconnect to rapidly transmit data worldwide. This laid the groundwork for the eventual creation of the Internet.

  • The summary emphasizes the innovation of packet switching, the geographic scale of the network, and how this early experiment enabled the future development of the Internet.

The internet was created through a collaborative effort between government agencies, researchers, and technology companies over several decades.

  • The origins of the internet trace back to the Advanced Research Projects Agency (ARPA, later DARPA) in the 1960s. DARPA funded research in packet switching networks that formed the basis for early networks like ARPANET.

  • ARPANET connected universities and research institutions and adopted an open-source ethic that enabled researchers to build on each other’s work. This led to the development of the TCP/IP protocols that define the internet.

  • Commercial use of TCP/IP expanded in the 1980s with services like Usenet and dial-up internet service providers. The World Wide Web was created at CERN in 1989.

  • In the 1990s, technologies like HTTP and HTML browsers emerged, enabling the web as we know it. Infrastructure expanded rapidly during the dot-com boom.

  • The internet was transitioned to being commercially managed and privately operated in the 1990s, moving away from government control. But government funding and collaboration between public agencies, researchers, and companies was essential to developing the fundamental technologies.

So in summary, the internet was the result of decades of work, built on previous innovations, and involved extensive collaboration between government, academia, and private companies. Its decentralized, open design has enabled the massive growth and innovation that followed.

  • The internet was originally created by the US military to connect computers over long distances. The goal was to bring computing power to the battlefield by linking mobile units with mainframe computers back home.

  • TCP/IP, developed by Cerf and Kahn, was the protocol that allowed different networks to communicate. It is the lingua franca of the internet.

  • The military wasn’t interested in the original idea, but embraced TCP/IP to connect its own networks. ARPANET switched to TCP/IP in 1983, creating what we now call the internet.

  • Civilian researchers began populating the early internet, using it for email, computing, and collaboration. The National Science Foundation then built out the backbone and regional networks to make the internet more widely available to researchers.

  • In the 1990s, the government privatized the internet, eliminating the NSFNET backbone and allowing commercial providers to take over. This coincided with the rise of the World Wide Web, which drove increased public enthusiasm and demand.

  • The government invented the internet, funded its early growth, then handed over the pipes to private companies without asking for anything in return. This privatization set the stage for the corporate-controlled internet we know today.

  • In the 1990s, there was a broad consensus to privatize the internet’s infrastructure, guided by free market ideology and deregulation under the Clinton administration.

  • The government didn’t literally hand over hardware to the private sector, but by killing the NSFNET backbone it ensured control passed to corporations. The TCP/IP protocols remained open, but the transition could have taken other forms.

  • Industry pushed the false choice between keeping the internet restricted or fully privatizing it. Large profits were at stake, so the government empowered industry to dictate privatization. This was a major but subtle victory.

  • Al Gore initially championed government oversight of the “information superhighway” but his stance changed as telecoms donated to Democrats. Public-private partnership disappeared from his rhetoric.

  • Alternatives like Senator Inouye’s “public lane” proposal went nowhere due to telecom lobbying. Critics warned of an “electronic shopping mall” but lacked mobilization.

  • The government committed to a market-dominated internet with industry self-regulation. Privatization wasn’t predetermined, but policies like expanding nonprofit networks were ignored. Industry shaped the debate.

The internet could have been established as a universal social right by supporting and expanding “free-nets” - nonprofit community networks that provided free local internet access, often through dial-up modems. Advocates tried to launch a federally-funded expansion of these networks, but telecom companies prevailed in privatizing the internet infrastructure instead. This privatization has enabled a few large companies to control the backbone networks and internet service provision, charge smaller networks for interconnection, kill competition, manipulate data flows, and make deals prioritizing certain content. The missed opportunity of establishing internet access as a universal right has allowed corporate interests to consolidate control over the pipes and plunder this essential infrastructure for profit. Reclaiming the internet as a public good will require dedicated efforts to enact pro-consumer regulations and policies.

  • The privatization of the internet in the 1990s was supposed to lead to better service, wider access, more investment, and lower prices. However, the evidence shows this has not been the case. Americans pay high prices for poor service compared to other countries.

  • ISPs act like slumlords - they charge high fees but underinvest in infrastructure, funneling profits to executives and shareholders instead. This leads to exclusion as well as exploitation.

  • Tens of millions of Americans lack broadband access, especially rural, low-income, and minority households. This digital divide was exacerbated during the pandemic.

  • Smartphones are not adequate substitutes for home broadband. Lack of proper internet access puts people at a disadvantage economically, educationally, and socially.

  • Internet access is a prerequisite for full participation in society. It is one of the material resources people need for individual self-rule and democracy.

  • Privatization allocates internet access solely based on profitability rather than treating it as a public good or universal right. This inherently limits democratic participation. The internet should be managed in the public interest.

  • In 1935, Chattanooga residents created a public electric utility because private monopolies refused to provide affordable electricity.

  • In 2010, this same public utility, the EPB, began offering ultra-fast 1 gigabit internet service called “the Gig.”

  • The Gig was enabled by a fiber optic network originally built for a smart grid system. Fiber to the home allows for extremely fast speeds.

  • The Gig offers some of the fastest residential internet speeds in the world at reasonable rates, with discounts for low-income residents.

  • There are over 900 communities in the U.S. served by public or cooperative broadband networks focused on universal access rather than profit.

  • Rural electric and telephone cooperatives have brought fiber-to-the-home gigabit internet to rural areas in North Dakota, outperforming private ISPs that take government subsidies but fail to invest in rural access.

  • Public and cooperative networks prioritize community benefits over profit maximization. They deliver better service at lower cost compared to private ISPs.

  • Community networks offer a model for reorganizing the internet around human need rather than profit. They are publicly or cooperatively owned, which enables them to operate differently than private firms.

  • Legal mechanisms like tax exemptions compel community networks to provide service at cost rather than accumulate profits. Democratic control by users is also mandated.

  • Community networks like those in Chattanooga, rural North Dakota, and Detroit’s Equitable Internet Initiative provide affordable connectivity to underserved areas and empower users to participate in governance.

  • By serving excluded communities and enabling community control, these networks challenge the privatized model of internet infrastructure. Corporate ISPs have tried to destroy them through lobbying and lawsuits.

  • What companies really fear is the demonstration effect - if people see community networks succeeding, they may demand this alternative model instead of privatized infrastructure.

  • Community networks help reopen political possibilities foreclosed by privatization in the 1990s. They make credible the idea that the internet can be organized around human need rather than profit.

  • Community networks represent an alternative to the current corporatized internet system. They are being built from the bottom up through creative community efforts.

  • To help community networks grow, they need to be defended from laws that restrict them, and extended through public funding and policy measures like requiring grant recipients to adopt democratic governance.

  • Public funding could support free or low-cost access, taking internet from a commodity to a public good.

  • Government and other anchor institutions could be required to purchase internet service from community networks, supporting their financial sustainability.

  • Community networks should be part of a broader ecosystem of democratic economic institutions like cooperatives, reinforcing shared values.

  • However, community networks also have limitations due to their localized nature. Larger-scale infrastructure and governance is still needed, like municipal networks and public options at higher levels. The aim is a “network of networks” that balances local and national.

  • Community networks face challenges in reaching the “interim internet” or core backbone controlled by large ISPs. These firms could block community networks’ upstream access.

  • To fully realize their potential, community networks need an alternative national backbone. This could be built through a Green New Deal electricity grid with fiber laid along it.

  • A federal agency or cooperative federation could manage the new backbone and expand it by acquiring unused fiber and stake in subsea cables.

  • Creating this alternative will require a movement of people taking disruptive action as was missing in the 1990s.

  • Concerns about costs can be addressed by pointing to the billions in subsidies already given to ISPs with little benefit. The money could be better spent on public options.

  • Concerns about government capability can be met by creating new public agencies focused solely on this mission or partnering with cooperatives. The goal is to avoid bureaucratic ossification.

  • Concerns about overreach can be alleviated by enacting strong privacy and free speech protections in law for the new public options. The point is to create real alternatives to corporate power, not replicate it.

In summary, building a democratic internet requires extending community networks to the core in a way that is cost-effective, capable, and rights-respecting. This will need a strong movement to overcome corporate opposition.

  • The build-out of publicly and cooperatively owned broadband networks should be prioritized over private networks. Ongoing operational support should be provided to keep service free or very low cost.

  • Additional revenue sources will be needed, such as taxes on tech giants like Google and Facebook or fees on large businesses to subsidize free/low-cost access for individuals.

  • Competition between private ISPs has limits in its ability to lower prices and increase access. Public networks can achieve broader access and give users more control.

  • Relying solely on consumer choice and market signals is a weak way for people to exercise power over their lives. Public broadband allows for more direct democratic control.

  • State surveillance and censorship can happen regardless of whether infrastructure is publicly or privately owned. What matters is how state power is used. Historical precedence like the post office shows publicly owned networks can prioritize privacy and freedom.

In summary, publicly owned broadband with democratic oversight is presented as a superior model to privatized networks for achieving universal, affordable access with greater user control and privacy protections.

  • In 1995, Pierre Omidyar created AuctionWeb, one of the first major online marketplaces. He was motivated by a desire to give individuals the power to be both producers and consumers online.

  • AuctionWeb quickly grew into a profitable business called eBay. It embodied some key features of what would later be called internet “platforms” - it facilitated interactions between users, relied on user contributions, and aimed to create an open and neutral online marketplace.

  • However, the term “platform” is imprecise and misleading. It obscures the power and control that companies like Google and Facebook wield over their digital spaces.

  • A better way to understand major internet companies is as infrastructure providers that shape our experience of the internet. They exercise sovereignty over their domains and actively structure how users interact.

  • The early spirit of eBay empowering individuals as both producers and consumers stands in contrast to later platform companies that mostly treat users as consumers from whom data and attention can be extracted.

  • The privatization of the internet began in the 1990s with the infrastructure layer (the “pipes”). By mid-1990s this was complete.

  • The next phase was privatizing the application layer, where users interact. This was the focus of the dot-com boom starting in 1995.

  • Hundreds of startups were founded and billions invested, but most were not profitable. There was no clear business model yet for monetizing internet activity.

  • The dot-com bubble burst in 2000-2001. Many saw it as a period of irrational exuberance and bad business decisions.

  • However, the real issue was that the internet had not yet been optimized for profit-making. The elements for capitalizing on user activity were not yet in place.

  • eBay offered a first glimpse of this new privatized model - it engaged users in its creation through their auction listings and marketplace interactions.

  • eBay represented a fusion of community and market - harnessing user activity itself as the basis for profit.

  • This built on the internet’s history as a social technology, pioneered by innovations like email that facilitated human connections.

  • Figures like Bill Gates saw the internet as the “ultimate market” - merging social relationships and market exchange. eBay operationalized this vision.

  • Pierre Omidyar founded AuctionWeb (later eBay) in 1995 with the goal of creating an online community marketplace.

  • The site was successful for three main reasons:

  1. Network effects - more users attracted more buyers/sellers and more listings, making it more valuable.

  2. Role as a middleman - eBay connected buyers and sellers but didn’t hold inventory, enabling it to profit while remaining lean.

  3. Sovereignty - eBay set the rules and policies governing user interactions, which was necessary to manage the community.

  • These factors allowed eBay to fuse community and market in a lucrative way. It turned a profit when many dot-coms failed.

  • Online platforms later built on these concepts, becoming like online shopping malls - privately owned public spaces where social and commercial interactions occurred.

  • A key difference was that platforms profited not just from facilitating transactions but from collecting data on all interactions that occurred within them.

  • This data collection and analysis became a primary focus, propelling the capitalist reorganization of the internet.

  • In the 1990s, the internet was growing rapidly with tens of millions joining each year. The amount of information online was exploding, making search and navigation challenging.

  • Page and Brin aimed to address this challenge with Google’s ranking algorithm that analyzed how websites linked to each other. This helped scale search to the rapidly expanding internet.

  • As Google grew into a business, it began logging user searches and clicks. This data was initially used to improve search relevance.

  • In 2002, Google began using this user data to target and rank ads, making its ad system far more effective. This marked a shift to monetizing user data.

  • Shoshana Zuboff argues this inaugurated a new form of “surveillance capitalism” where user data is extracted and commodified.

  • Google pioneered an online advertising model that relied on extensive tracking of user behavior. This fueled its profits and dominance.

  • Other sites like Facebook also capitalized on user-generated content and data. Together, Google and Facebook now control over 50% of the digital ad market.

  • Amazon and eBay emerged from the dot-com era as e-commerce pioneers, but had different business models. eBay was an online marketplace that didn’t handle fulfillment, while Amazon took on the costs and complexities of warehousing and shipping products directly to customers.

  • In the early years, eBay’s model looked smarter, with higher margins and profits. But after the dot-com crash, Amazon adapted to become more eBay-like by opening its platform to third-party sellers through Marketplace.

  • Letting third-party sellers list on Amazon and compete on price was controversial internally but quickly paid off, driving rapid growth in sales and Amazon’s first profitable quarter in 2002.

  • Over time, Amazon aggressively courted third-party sellers away from eBay. Combined with missteps by eBay, this allowed Amazon to overtake its rival and become the dominant online marketplace.

  • Like Google and Facebook, Amazon’s rise was powered by its evolution into a “platform” - an online mall that sold not just its own goods but provided space for countless vendors. This marketplace model proved hugely lucrative.

  • By 2007, Amazon had become the dominant player in e-commerce, with $14.8 billion in sales. E-commerce went mainstream in the 2000s, and Amazon was well positioned to capitalize on this with its streamlined storefront and logistics network.

  • Building a third-party marketplace enabled Amazon to expand its offerings by opening up its site to millions of small and mid-sized sellers. This generated substantial revenue for Amazon through fees and services.

  • Third-party sellers also provided valuable data that Amazon could use to launch copycat products and conduct market research. The marketplace served as a “petri dish” for Amazon.

  • As the internet commercialized, it required more complex machinery and systems. The “cloud” emerged - a network of data centers providing on-demand computing resources over the internet.

  • Though an old idea, Amazon was key in realizing the concept of cloud computing. A team in South Africa built Elastic Compute Cloud (EC2), laying the groundwork for on-demand computing services that would be highly profitable for Amazon.

  • The cloud industrialized parts of the internet, providing the infrastructure for the data processing and software needed in the commercial web environment.

After the dot-com crash, Amazon began developing a cloud computing service called Amazon Web Services (AWS) to better manage its infrastructure needs. A team worked on this in Cape Town, away from Seattle and Bezos’s micromanaging. They used virtualization technology to allow customers to easily rent computing resources like virtual servers on demand. AWS launched in 2006 with EC2 for computing and S3 for storage. It enabled smaller companies to scale capabilities like data storage without massive infrastructure investment.

AWS arrived right as internet companies were amassing more data and needing to monetize it. The “data imperative” led organizations to collect as much data as possible to try to utilize artificial intelligence. AWS provided the tools for this data processing and AI in the cloud.

In the 2010s, the internet went mobile with smartphones. More objects also got networked connections in the “Internet of Things.” The internet became ubiquitous instead of tethered. “Smartness” referred to an object’s ability to connect to the cloud and exchange data. The network became the universal everywhere computer. Even basic objects like coolers got sensors, cameras, and screens to watch customers, analyze data, and target ads, demonstrating this new networked intelligence.

  • Smart coolers with facial recognition and iris tracking were envisioned but never implemented. However, the story shows a broader trend of making everyday objects “smart” and connected to put more aspects of our lives online for data collection and targeted advertising.

  • This “internet of things” recalls the original vision for the internet as a network with mobile nodes, putting the computers where the people are. DARPA saw potential for command and control via the internet.

  • Uber exemplifies algorithmic management via the internet - software directing drivers what to do. This enables coordination of millions of gig workers without managers. Classifying workers as contractors rather than employees increases profits.

  • Since the 1990s, corporate networks enabled outsourcing and offshoring of work. The internet has accelerated this “fissured workplace” by making it easier to parcel out work to contractors remotely.

  • Platforms like Uber combine the internet’s ability to distribute work globally with new techniques of algorithmic management for heightened control and profits. Workers are remotely monitored and directed in detail but treated as independent contractors.

  • The result is a pronounced power asymmetry between workers and the platforms, enabled by the architecture of the internet. This recalls critiques of early ARPANET as facilitating centralized control. The internet’s flexibility can enable liberation but also new forms of control.

  • In the 1970s, the shipping container enabled companies to outsource manufacturing operations overseas where goods could be produced more cheaply. This eliminated many blue-collar jobs.

  • The internet has similarly eliminated many white-collar jobs by creating a “human cloud” of freelancers and subcontractors that companies can tap into on demand. Tech companies use these workers to label data for machine learning algorithms.

  • If the internet just moved work around it wouldn’t be that useful. The key is retaining control over the workflow fragments. Companies use the same networks to distribute and monitor work. Uber represents this trend taken to the extreme.

  • Uber loses billions of dollars but is still valued at tens of billions due to the data it produces. This data attracts investor capital by holding out the promise of future optimization and profits, whether or not those materialize.

  • Even if Uber never becomes profitable, early investors and shareholders have already made huge returns by cashing out. Profits can be made from speculation rather than production.

  • Data manufacturing and speculation have become central to internet privatization, representing a new “baroque” phase. The internet has become saturated as a source of profits but unprofitable firms like Uber find new speculative value in data.

Here is a summary of the key points about successful internet companies:

  • Successful internet companies like Google, Facebook, and Amazon have accrued huge profits and minted billionaires, but this has coincided with stagnating wages and rising inequality.

  • These companies make money in different ways - by selling access like telecoms, or by monetizing data and activity like online platforms.

  • Online platforms especially rely on a mix of technology, politics, legal arrangements, and investor financing. These elements interact to enable their business models.

  • For example, Uber leverages its tech and data trove, but also uses lobbying to change regulations in its favor and investor money to subsidize cheap fares.

  • So while online platforms share some traits, their specifics and effects differ based on their embeddedness in wider institutional forces.

  • However, a common result is that these online platforms tend to increase economic inequality, even if in varied ways. They concentrate wealth and power.

  • Uber and other “gig economy” companies have redistributed risk and reward in the labor market. They push risks like income instability onto workers while pulling rewards like profits to executives and investors.

  • This “predatory inclusion” brings marginalized groups like minorities and women into the workforce but on exploitative terms with low pay and few benefits.

  • A similar dynamic exists in online spaces, where minorities and women experience new forms of exclusion and bias. What scholars call “cybertyping” adapts old prejudices to new digital formats.

  • As tech platforms have centralized control, cybertyping has evolved. Oppression is increasingly embedded in and amplified by automated systems like algorithms and AI.

  • Overall, online malls redistribute risk and reward by absorbing excluded groups into their systems in ways that entrench their vulnerability. Rewards flow to a small technology elite while precarity spreads.

  • Oppression has become algorithmic in the age of online malls like Google. Information used to be organized mostly by humans, now it is organized by algorithms.

  • This can lead to biased and stereotypical results, like when searching for “black girls” returned porn sites. It’s a form of “predatory inclusion” - including marginalized groups but in harmful, stereotypical ways.

  • Online malls also enable far-right groups to reach wider audiences than before. Conservatives have always been “innovation opportunists” with new media technologies.

  • Social media in particular has accelerated far-right politics. But the popular narrative about “filter bubbles” causing polarization is overly simplistic.

  • Polarization is complex and social media use doesn’t neatly map onto it. There is false equivalence between the Left and Right - conservative movements have benefited far more from weaponizing social media.

  • Social media platforms optimize for engagement, which can advantage inflammatory content. But this doesn’t directly cause radicalization. The roots are in longer-standing right-wing projects.

The Right has far more resources to exploit the shift to digital media than the Left, including deep-pocketed donors, sophisticated media operations, and control of the Republican Party. Right-wing pundits and false information, not foreign operatives, are the main purveyors of misinformation online. The right-wing media ecosystem allows propaganda to spread through outlets like talk radio, cable TV, and websites. This “trading up the chain” tactic lends credibility to false narratives by getting them repeated in more mainstream outlets. However, internet users are not simply brainwashed - they actively argue over claims. Right-wing propagandists employ creative tactics to increase engagement and spread narratives on platforms like Facebook and YouTube.

Though social media algorithms can favor polarizing content, the companies also face pressure to crack down when violent events like the Capitol riot happen. They aim to strike a balance that allows right-wing engagement up to the point of alienating advertisers. Republicans accuse the companies of anti-conservative bias, pressuring them to be even more hospitable to right-wing content. While profitable in the short term, an overinvestment in the Right could backfire if Democrats gain power and regulate the companies. The symbiotic relationship between social media and right-wing networks is complex, as are the techniques propagandists use to exploit the system.

The passage discusses different perspectives on reforming the internet and online technology companies. The key points are:

  • There are two main strains of internet reform: 1) Writing new rules about how companies can behave, 2) Reducing the market power of big technology firms, advocated by the “New Brandeisians.”

  • The New Brandeisians want to make markets more competitive, believing this will distribute wealth more widely and reduce corporate political influence. They support breaking up large companies or constraining “natural monopolies” through regulation.

  • Each approach has merits - the rule-makers rightly say tech companies are under-regulated, while the New Brandeisians recognize the current ownership structure is constructed through public policy.

  • However, more competition may not solve problems like the focus on maximizing user engagement. Competitive pressure drives questionable practices, and monopolies can afford moderation efforts that competitors cannot.

  • Regulation also has downsides like regulatory capture and erecting barriers to entry. There are no easy solutions, but recognizing the privatized nature of the internet is an important first step.

Large corporations can bear more regulation, but real change requires going beyond regulating markets to deprivatizing them. Online platforms are built for profit-making, which generates inequality and social harms even with regulation. Mark Zuckerberg’s power over Facebook shows the inequality of private ownership.

To move past privatized online platforms, we can’t just replace them with similar public or cooperative clones. We need new digital architectures that allow collective governance. Angela Davis’ idea of police/prison abolition provides a model - building alternative institutions rather than just reforming existing ones.

This requires imagination and experimentation with new ways of structuring online spaces. Instead of centralized social media giants, we could have many self-governing communities of different sizes and purposes that users can participate in meaningfully. They can be decentralized but interconnected. Public protocols and shared data infrastructure can tie these communities together while keeping power distributed. We need to explore many options, from municipal networks to cooperatives to new hybrid forms. The goal is a pluralistic digital sphere beyond privatized enclosures.

  • The internet allows networks to interconnect through open, non-proprietary protocols. This allows an open network where any participant can join if they follow the rules.

  • Mastodon applies this decentralized model to social media. It is part of the “Fediverse”, which includes other decentralized social media projects.

  • Decentralized social media preserves the benefits of network effects without centralized control. It resembles email, where different services can interoperate through common protocols.

  • Privatized social media has walled off sections of the online world into closed, proprietary spaces governed by secret algorithms. Decentralization can break down these walls.

  • To make decentralized social media more robust requires public investment, such as through public libraries and public media. Libraries can provide infrastructure and accountability. Public media can provide higher quality information content.

  • Reforming social media is about investing in forms of care work that are currently undervalued, like content moderation.

  • Decentralized social media also allows for better moderation, like quarantining fascist communities while preserving free speech.

  • More broadly, a “public option” for cloud computing and cooperative platforms for services like taxis can provide alternatives to privatized online spaces. Public policy tools like grants and regulations can support these efforts.

  • The London Greater Council in the 1980s established Technology Networks - prototyping workshops where people could access tools and expertise to build things based on community needs rather than profit. This model could be replicated today to democratize internet technology design.

  • Blurring the line between creators and users of technology, so that expertise is not just technical but based on lived experience, is key to creating an internet ruled by the people. The disability rights slogan “Nothing about us without us” encapsulates this.

  • New cooperatively owned online services need to be embedded in supportive institutional relationships to be sustainable - for example, through public procurement contracts, partnerships with other coops, government support.

  • Adversarial interoperability (requiring tech giants to adopt open protocols enabling interconnection) and anti-monopoly legislation are important offensives against the dominance of major online platforms.

  • Ultimately, radically reimagining the internet requires expanding who participates in shaping it, blurring the line between creators and users. The goal is an internet built from the bottom up based on human needs.

  • Facebook makes money from advertising revenue generated by users interacting on its platform. Decentralized alternatives like Mastodon threaten this business model by giving people other options. If Facebook users could easily interact with Mastodon users, more people might join Mastodon, shrinking Facebook’s user base and revenue.

  • Remaking the internet requires political organizing, not just technical solutions. The early internet pioneers in 1980s London built technologies but also campaigned for social change like the “Right to Warmth” initiative.

  • To dismantle today’s online monopolies, we need social movements that target the foundations of privatization and reimagine the internet’s future, inspired by groups like the Luddites. This means creativity to envision alternatives combined with enough disruption to force change.

  • Strategies for change must be tailored to different layers. Community-owned networks can provide affordable local infrastructure. Dismantling online monopolies requires anti-trust action plus building cooperative alternatives not driven by profit-seeking.

  • The goal is an internet organized around democratic principles and human needs, not corporate interests. This will take a diversity of approaches, public investment, and determination to reshape both technology and society.

  • In November 1977, an experiment was conducted that involved sending data packets across multiple networks, including the early internet (ARPANET), demonstrating an early form of internetworking. This took place in Silicon Valley using a van equipped with radio and computer equipment.

  • The origins of the internet lie in the Advanced Research Projects Agency (ARPA, later DARPA), formed in 1958 in response to Sputnik. ARPA funded the development of computer networking, leading to the creation of ARPANET in 1969, the precursor to the internet.

  • The internet was developed through government funding, largely shielded from commercial pressures. It was an unlikely creation, as the idea went against dominant trends in computing at the time.

  • Key innovations like packet switching (breaking data into packets routed over a distributed network) enabled the internet’s functionality and were developed through government-funded research. Private telecom firms like AT&T initially showed no interest.

  • The 1977 experiment demonstrated early global internetworking capabilities and helped popularize the technology. It highlighted the non-commercial, experimental roots of the internet before its privatization and commercialization.

  • ARPANET was an early packet-switching network developed by the U.S. Department of Defense’s Advanced Research Projects Agency (DARPA) in the late 1960s and early 1970s.

  • The contractor building the Interface Message Processors (IMPs) for ARPANET, Bolt, Beranek and Newman (BBN), initially refused to share the source code for the IMP software. This caused conflict within the ARPANET community.

  • DARPA ultimately forced BBN to share the IMP source code with the other contractors. This allowed the networking technology to spread.

  • In the 1970s, DARPA began early research into interconnecting networks, which led to the development of the TCP/IP protocol by Vint Cerf and Bob Kahn. This allowed different networks to communicate.

  • TCP/IP was seen as a way for the military to interconnect its various networks. The early testing and growth of TCP/IP on ARPANET and other networks was funded by DARPA and other government agencies.

  • The government played a key role in the development and spread of the early internet, including subsidizing the creation of regional networks and implementation of TCP/IP. The government saw the potential benefits of networking for research and military communication.

  • In the 1980s and early 1990s, privatization of the internet backbone was planned and encouraged by the government. The National Science Foundation transitioned the backbone to commercial providers in 1995.

  • The internet’s backbone network was originally built by the government and operated by nonprofits and universities. But in the 1990s, it was rapidly privatized and handed over to large telecom companies.

  • This privatization happened with little public discussion or oversight, driven by an emerging political consensus that the internet should be turned over to “the market.” Critics argued this would undermine the public interest nature of the early internet.

  • The major backbone providers that emerged quickly consolidated, leaving just a handful of large companies in control. They also kept their interconnection agreements secret.

  • On the last-mile side, decades of deregulation allowed a few large broadband providers to dominate most markets with little competition. This was cemented by the Telecommunications Act of 1996.

  • As a result, most Americans today have few choices for home broadband. Large providers can use their gatekeeper power to prioritize their own content, charge access fees, and undermine net neutrality principles.

  • Recent interconnection deals between companies like Netflix and Comcast illustrate how large players can find ways to benefit each other, while smaller networks and startups get disadvantaged.

  • Overall, privatization and deregulation have allowed a few powerful companies to exert significant control over the internet’s infrastructure, often in ways that serve private interests rather than the public good.

Here is a summary of the key points from the excerpts:

  • The internet in the US is controlled by a small number of large corporations like Comcast and AT&T that have local monopolies and engage in anti-competitive practices. This leads to high costs, slow speeds, and poor customer satisfaction compared to other countries.

  • There are large disparities in internet access, with rural, low-income, and minority communities being disproportionately disconnected. This “digital divide” was exacerbated during the COVID-19 pandemic when internet access became even more essential.

  • In response, some communities like Chattanooga, Tennessee have created locally-run, publicly-owned broadband networks called “community networks.” These often provide cheaper, faster, more reliable internet.

  • Community networks treat the internet as a public utility rather than a private commodity. They are run democratically by the community for the community.

  • However, large telecom companies lobby to restrict community networks, seeing them as threats to their monopolies. Defending community networks will require political organizing and overcoming legal barriers in many states.

  • Overall, the excerpt argues that internet access should be recognized as a public good necessary for freedom and democracy. It advocates community-run networks as a way to make the internet more equitable and accountable to citizens rather than corporations.

  • In 1995, Pierre Omidyar founded AuctionWeb, which later became eBay, as an online marketplace aiming to create a “perfect market.” Omidyar saw the internet as a way to enable unfettered free trade between individuals.

  • eBay grew quickly as a platform for individuals to buy and sell items directly with each other, bypassing traditional retailers. By charging listing fees and a percentage of each sale, eBay turned a profit within its first year.

  • However, eBay’s model also enabled fraud, counterfeit goods, and abusive behavior among users. Omidyar initially took a hands-off, libertarian approach, trusting users to police themselves through eBay’s feedback system.

  • As eBay grew into a multibillion dollar company, it was forced to exert more control over transactions, mediate disputes, and ban problematic users, moving away from Omidyar’s original perfect market vision.

  • eBay demonstrates how online marketplaces require interventions to function smoothly, belying techno-libertarian views of the internet as an unfettered free market. Private governance and regulation become necessary at internet scale.

Here is a summary of the key points from the excerpt:

  • In 1993, the Clinton administration envisioned an “information superhighway” where people could shop from home. This was likened to a virtual shopping mall.

  • The first modern shopping malls were developed in the post-WWII years by architect Victor Gruen, who saw them as quasi-public spaces that would benefit communities.

  • Today, tech platforms function as online malls that enable rent-seeking, extracting fees from the transactions they facilitate.

  • In 1998, Google founders Larry Page and Sergey Brin published a paper outlining an improved web search algorithm called PageRank.

  • Early search engines like AltaVista struggled with too much data, producing spammy results. PageRank helped Google provide more relevant results.

  • As Google’s search logs grew, analyzing user data became key to improving its algorithms and targeting ads. This marked a shift toward data extraction and surveillance.

  • Google began selling targeted text ads in 1999, marking the start of its platform business model based on selling ads optimized via user data.

  • Kozmo, Webvan, and were early e-commerce companies that went bankrupt due to logistical challenges with delivery and fulfillment.

  • In contrast, Amazon was able to survive the dot-com crash. A key move was launching Amazon Marketplace in 2000, which allowed third-party sellers to list products on Amazon’s site. This greatly expanded Amazon’s product selection without the company having to stock inventory.

  • By 2009, the value of goods sold on Amazon Marketplace was more than double what was sold directly by Amazon. Today, over half of paid units sold on Amazon are from third-party sellers.

  • Amazon collects extensive data on third-party sellers and has used this to launch its own competing products, a controversial practice.

  • The on-demand model of cloud computing has roots going back to early ideas about computing utilities in the 1960s. The modern cloud computing industry emerged in the mid-2000s with Amazon Web Services.

  • AWS allowed companies to rent computing infrastructure on demand rather than having to buy and maintain their own servers. This allowed startups to scale quickly and cheaply.

  • AWS has become a highly profitable engine of growth for Amazon. Cloud services now make up the majority of Amazon’s operating income.

Here is a summary of the key points from the excerpt about Compaq and NetCentric:

  • In 2004, an Amazon employee named Chris Pinkham was tasked with finding a way to simplify Amazon’s chaotic internal infrastructure. He came up with the idea of selling basic computing services over the internet.

  • Pinkham and colleague Benjamin Brown set up a team in Cape Town, South Africa to develop this idea. The location was chosen in part due to cheap electricity costs.

  • The Cape Town team built a virtualization service that allowed Amazon to sell excess computing capacity from its data centers. This service launched in 2006 as Amazon Elastic Compute Cloud (EC2).

  • EC2 enabled companies to rent servers and storage from Amazon rather than operate their own data centers. This kickstarted the market for cloud computing services.

  • The on-demand nature of EC2 allowed companies to quickly scale up or down based on computing needs. This supported new types of applications and businesses built on the cloud.

  • Other Amazon cloud services like S3 storage soon followed. Together, these formed the origins of Amazon Web Services, which pioneered the public cloud computing model.

This passage discusses how tech companies like Uber have adopted a model of “predatory inclusion” that appears inclusive on the surface but in reality exploits vulnerable workers.

The key points are:

  • Uber undermined regulations and worker protections that taxi drivers had previously won through collective action, leading to deteriorating conditions and poverty wages for rideshare drivers.

  • Inspired by Uber’s success, companies across industries are classifying more workers as independent contractors to avoid providing benefits and skirt labor laws.

  • Tech companies like Google and Facebook rely heavily on contract/temp workers who perform critical functions but receive low pay and minimal benefits compared to full-time employees.

  • The inclusion of marginalized groups in tech platforms and gig work masks the extractive nature of these business models.

  • From early on, the internet has replicated existing racial biases and enabled new forms of racial discrimination, despite claims it would be an egalitarian, enlightened space.

  • Search engine algorithms like Google’s amplify stereotypes by surfacing problematic associations in response to searches about marginalized groups.

In summary, the “predatory inclusion” of vulnerable workers and racial minorities in tech platforms serves to benefit tech companies while exploiting these groups and concealing deeper inequities.

Here is a summary of the key points from Noble’s article “orithmic hypervisibility in Noble, “Google Search: Hyper-visibility as a Means of Rendering Black Women and Girls Invisible,” InVisible Culture 19 (October 2013)“:

  • Noble argues that Google’s algorithmic search results contribute to the hypervisibility of stereotypical and sexualized images of Black women and girls, while simultaneously rendering them invisible as full human subjects.

  • She provides examples of racist and sexist search suggestions and image results that emerge from neutral search terms related to Black girls. These reflect embedded societal biases.

  • The invisibility occurs because the algorithmic assemblage of words and images reduces Black women and girls to stereotypes, preventing meaningful engagement with their full humanity.

  • Noble situates this within historical contexts of technology being used to marginalize Black women, from slave catalogs to CCTV monitoring. Algorithmic search continues this marginalization through hypervisibility of stereotypes.

  • She argues that tech companies must take responsibility for such biased outcomes and build more equitable algorithms. This entails acknowledging the politics embedded in tech design.

In summary, Noble highlights how Google Search’s algorithmic logic contributes to the harmful hypervisibility of racist and sexist stereotypes about Black women and girls, ultimately working to keep them invisible as fully realized human beings. She calls for greater accountability in algorithmic design.

  • The “techlash” against big tech companies like Facebook gained steam in 2018, with more public skepticism about their impact.

  • Two main responses emerged: breaking up or tightly regulating tech giants, and developing alternative online institutions like cooperatives and nonprofits.

  • The “New Brandeisians” advocate antitrust action to curb the power of tech monopolies. This view is gaining support, seen in government investigations and lawsuits against companies like Facebook.

  • Others argue that turning tech firms into tightly regulated public utilities could better serve the public interest than breaking them up.

  • More radical proposals involve developing alternative online spaces like decentralized social networks and platform cooperatives owned by users. These aim to serve human needs rather than maximize profits.

  • Additional ideas include funding independent, nonprofit journalism; treating content moderation as care work; and creating democratic data governance models.

  • Efforts to build a more humane internet require public funding and spaces for experimentation, learning from past initiatives like the Greater London Council’s Technology Networks in the 1980s.

  • The goal is an internet that serves human needs and embodies cooperative design principles rather than concentrating power in large corporations.

Here is a summary of the key points from the referenced page:

  • The early internet was built through public funding, especially from the U.S. military and research institutions. It was designed as an open network for sharing information.

  • In the 1990s, the internet began to be privatized and commercialized. Internet service providers (ISPs) emerged, charging for internet access. E-commerce sites like Amazon and eBay also arose.

  • The bubble in the early 2000s led to a shakeout of internet companies. Survivors like Google, Facebook, and Amazon came to dominate through their platforms and accumulation of user data.

  • Today’s internet is largely owned and controlled by private companies. A few tech giants wield enormous economic and political power. Their business models often exploit users’ data.

  • This raises concerns about privacy, monopoly power, political polarization, misinformation, and inequality of access. Some argue the internet should be treated more as a public utility than a private business.

  • There are efforts to create alternative community-owned internet networks. But challenges remain in making internet access more equitable and keeping tech power in check. The future form of the internet remains contested.

Here is a summary of the key points about the history and development of the internet:

  • The internet originated from military research and university experiments in networking in the 1960s and 1970s, especially ARPANET created by the Defense Advanced Research Projects Agency (DARPA).

  • The TCP/IP protocol, developed by researchers like Robert Kahn and Vint Cerf in the 1970s, allowed different networks to interconnect into the internet.

  • The National Science Foundation (NSF) funded the expansion of the internet backbone and connections to universities in the 1980s through NSFNET.

  • The World Wide Web, created by Tim Berners-Lee at CERN in 1989-90, made the internet accessible through hypertext and browsers.

  • Commercialization began in the 1990s, with companies like AOL, Netscape, Yahoo, eBay, and Amazon emerging. The NSF lifted restrictions on commercial activity on the internet.

  • Internet infrastructure was privatized in the 1990s, with telecom companies taking over the backbone and access provision. Deregulation enabled this.

  • The rise of search engines like Google, social media like Facebook, and smartphones drove massive growth after 2000.

  • More recently, there are debates about concentration of power, privacy, and misinformation on social media platforms. Some advocate decentralization and public alternatives.

Author Photo

About Matheus Puppe