There is a strange experience that has become increasingly familiar in our digital lives: a technology we once liked slowly becomes worse. A social media platform that once showed us posts from people we chose to follow becomes an endless stream of advertisements, sponsored content and algorithmic recommendations. A service that was once free begins locking useful features behind subscriptions. A device that once lasted for decades becomes increasingly difficult to repair. Software we purchased outright becomes something we must continuously rent. Sometimes, we even discover that the digital products we thought we owned were never really ours at all.
It is tempting to see each of these frustrations as a separate problem. But veteran software engineer and open-source advocate Paul Fenwick argues that they are connected by something much bigger: the economic systems shaping how our technologies are designed. There is even a word for what is happening: enshittification. Speaking on NENA’s Voices of the New Economy podcast, Paul explores why technology seems to be getting worse, how surveillance capitalism and concentrated corporate power contribute to the problem, and what alternatives might allow us to reclaim greater control over our digital lives.
What is enshittification?
The term enshittification was coined by technology writer Cory Doctorow to describe the process through which digital platforms progressively deteriorate as companies seek to extract more value from the different groups that depend upon them. Many platforms begin by being extremely good to their users. A new social network wants people to join, so it creates a useful service with relatively few advertisements and restrictions. A marketplace makes it easy for buyers to find sellers. A streaming platform offers an enormous library at an attractive price.
Once enough people have joined, however, leaving becomes harder. Our friends are on the platform. Our photographs are stored there. Our businesses rely on it to reach customers. We have built playlists, profiles, audiences and communities inside systems controlled by somebody else. That dependence creates an opportunity for extraction. Platforms can introduce more advertising, increase fees, reduce organic reach, collect more data, manipulate recommendations or make previously available features premium products.
The important point is that the technology hasn’t simply become worse because its designers forgot how to make good products. In many cases, the incentives surrounding the technology have changed. Once growth and market dominance have been achieved, the question shifts from how do we attract users? towards how much value can we extract from the users we already have?
Technology doesn’t exist outside the economy
This points towards a much bigger question than whether our favourite app has become annoying: what is technology actually designed to do? We often talk about technology as though it is neutral. A piece of software or hardware is simply a tool, and what matters is how humans choose to use it. But technologies are also created inside particular social and economic systems. When the success of a technology company is measured primarily through growth, profitability, market share and shareholder returns, its products will inevitably be shaped by those priorities.
This becomes particularly significant when companies achieve enormous market power. Competition theoretically creates an incentive for companies to keep improving their products: if customers dislike what one company offers, they can move somewhere else. Digital markets, however, often work differently. Network effects mean a social platform becomes more useful as more people join it. Digital ecosystems can make it difficult to move data between competing services. Devices may only work properly with software produced by the same company. Businesses can become dependent upon platforms for access to customers.
Eventually, choosing something else isn’t necessarily much of a choice. When only a handful of corporations control the infrastructure through which enormous parts of our lives operate, those companies acquire extraordinary power over the conditions under which we participate. Enshittification is therefore not simply a design problem. It is also a problem of ownership, power and the economic incentives embedded within our technological systems.
When you aren’t the customer, you might be the product
Perhaps nowhere is this clearer than in the rise of surveillance capitalism. Many of the digital services we use appear to be free, but providing sophisticated global digital infrastructure obviously isn’t. Instead, many platforms have developed business models based around collecting information about us: where we go, what we search for, what we click, what we purchase, who we communicate with, how long we look at something and what captures our attention. These enormous datasets can be used to predict behaviour, personalise advertising and optimise platforms around keeping us engaged for longer.
This changes the relationship between users and technology. A platform designed primarily to help us communicate would have an incentive to make communication better. A platform whose revenue depends upon capturing and monetising attention has an incentive to maximise the amount of attention it captures. Those goals aren’t necessarily the same.
The problem therefore becomes bigger than privacy. It raises questions about who owns our digital infrastructure, who gets to establish the rules governing our online spaces and who benefits economically from the enormous amount of human activity taking place within them. If our social lives, creative work, businesses and access to information increasingly depend upon privately controlled digital spaces, then questions about technology are also questions about economic democracy.
Planned obsolescence takes the same logic into the physical world
The same economic incentives shaping software also influence the physical devices we use. For much of human history, repairing things was ordinary. Clothes were mended. Furniture was restored. Tools were maintained. Appliances were repaired rather than immediately replaced. Yet many contemporary products have become increasingly difficult to repair. Components can be glued together rather than screwed into place, replacement parts may be unavailable, proprietary software can prevent third-party components from functioning, and repair manuals may be inaccessible. In some cases, replacing an entire product can seem easier or even cheaper than repairing one component.
For manufacturers operating within an economic system dependent upon continually expanding sales, this makes a certain kind of economic sense. A product that lasts for thirty years only needs to be sold once; a product replaced every three years creates ten purchasing opportunities over the same period. But what makes sense for an individual company’s revenue does not necessarily make sense for society or the planet. Every replacement device requires materials to be extracted, processed, manufactured and transported, while discarded products contribute to growing waste streams.
The humble act of repairing something therefore becomes surprisingly political. Paul’s own work as a volunteer repairer at repair cafés reflects another possible relationship with our possessions. Repair cafés and community repair initiatives do more than keep objects out of landfill. They preserve practical knowledge, reduce dependence on manufacturers and rebuild cultures in which maintaining what we already have becomes normal again. They also remind us that technological knowledge doesn’t have to remain concentrated within corporations: skills can be shared between neighbours and communities.
Generative AI adds another layer
The rapid expansion of generative artificial intelligence makes these questions even more urgent. AI is often presented primarily as a story of technological capability: what can these systems do today, and what might they be capable of tomorrow? But equally important questions concern ownership and power. Who controls the infrastructure? Where does the data used to build these systems come from? What happens to the work of artists, writers, programmers and other creators whose output contributes to the digital ecosystems from which AI systems learn? What information are users giving companies when they interact with AI tools? And what happens if an increasingly large proportion of the internet becomes mediated through a small number of enormously powerful technology companies?
The problem isn’t necessarily technology itself. The deeper question is what happens when extraordinarily powerful technologies develop inside economic structures already characterised by surveillance, monopoly and relentless pressure for growth. If we focus only on what AI can do without asking who owns it, who benefits from it and what incentives determine how it develops, we risk reproducing many of the same problems that have shaped earlier generations of digital technology.
What would technology look like if we treated it as a commons?
Fortunately, the corporate technology ecosystem isn’t the only model available. One alternative already has a long history: Free and Open Source Software (FOSS). Open-source software makes its underlying code available for people to inspect, modify and redistribute according to its licence. Instead of users being entirely dependent upon a single company to determine what software does, communities can participate in maintaining and developing it.
Projects such as Mastodon offer alternatives to corporate social media, while tools such as Krita and Jellyfin demonstrate how creative software and media infrastructure can operate outside conventional proprietary models. None of these solutions individually dismantles surveillance capitalism, but they demonstrate an important principle: digital infrastructure can be built differently.
Perhaps the most interesting possibilities emerge when we combine open-source technologies with another idea found throughout new economy movements: the commons. Rather than imagining our digital lives as a choice between individual consumers and giant corporations, we can ask what technologies might look like if communities collectively owned, governed and maintained more of the infrastructure they depend upon. Community-run servers, cooperatively owned platforms, shared digital infrastructure, open protocols that allow people to move between services, local repair networks and libraries of tools and equipment all point towards different relationships between people and technology. These approaches shift us from being passive consumers towards becoming participants in shaping the technological systems we use.
Reclaiming our technological agency
There is something surprisingly hopeful about recognising that technology is getting worse because of particular economic incentives. If technology were simply following an inevitable trajectory, there would be very little we could do. But technology is designed, business models are designed, markets are regulated, ownership structures are created, and communities establish norms about what they will and won’t accept. That means all of them can potentially be designed differently.
We can support the right to repair, choose open-source tools where practical, resist unnecessary surveillance, build cooperatives and community infrastructure, keep older devices functioning, share skills with our neighbours and demand stronger limits on monopoly power and corporate extraction. None of these actions alone creates a democratic technological future, but together they begin changing our relationship with technology from one of dependence towards one of agency.
The question facing us isn’t simply whether we want more or less technology. It is what kinds of technology we want, who should control them, and what purposes they should ultimately serve. Technology is never created in isolation from society. If we want technologies that are durable, democratic, repairable, privacy-respecting and genuinely designed to improve human life, we also need to build the economic and social systems capable of producing them.
Perhaps pushing back against enshittification begins with remembering something deceptively simple: the technologies we depend upon don’t have to work this way.
