It’s a familiar pattern. First there’s an economic crisis. Then comes an enormous restructuring of capital — and with it a restructuring of labor — throwing past certainties into doubt. Old industries, companies, and occupations disappear and new ones emerge, enabled by new technologies. As people struggle to find a way to describe the seismic upheavals in the economic and social landscape, a linguistic mist arises, muddling the features of this new landscape, blurring attempts to analyze and map it.

This time around, among the first commentators on the scene were techno-utopians, offering terms like “sharing economy” and “peer-to-peer networking” to conjure up a vision of a positive new development that, they claim, prefigures what a cooperative, post-capitalist society might look like. This is a society in which the Internet allows services to be freely shared between those who need them and those who can supply them, eliminating the capitalist intermediary. With 3-D printing, they even speculate, we can cut out factories too, allowing goods to be produced when they are needed.

The notion that people are happy to offer their services altruistically to strangers is picked up in another set of discourses in which the “wisdom of crowds” is tapped to “crowdsource” or “cloudsource” solutions to intractable problems.

Perhaps the clue here is in the word “source,” which evokes the corporate practices that took root in the 1990s, such as outsourcing (and variants such as global sourcing, insourcing, or backsourcing and their near relations, offshoring and nearshoring).

When terms like crowdsourcing are deployed, they’re used not to describe a voluntary gift economy but one in which online platforms are used to access a global pool of workers, on a just-in-time basis. In phrases like “liquid labor” and “workforce on demand,” this logic is even more apparent.

This on-tap workforce (the human cloud or crowd) is sometimes described pessimistically as “artificial artificial intelligence.” Its labor, vital for the myriad tasks that keep the internet humming, is often analogized to the Mechanical Turk, the nineteenth-century “automated” chess board that actually concealed a human operator.

Whether it is checking the accuracy of Google ratings, decoding street numbers for Google maps, deciding which horrific images of child sexual abuse or beheadings should be pulled from social media, or matching offers with requests on job search platforms, many of the activities most users imagine to be performed automatically by algorithms turn out in reality to be carried out by human “clickworkers.”

Optimists hyping the creativity and autonomy associated with freelance click-work call it the “gig economy”; more neutral commenters might describe a “platform economy” or “mesh economy.”

Whatever we call it, most of us have now accumulated enough anecdotes, if not firsthand experience, to have formed a mental image of what it is like to be part of the contemporary on-demand workforce. There are horror stories about Amazon Mechanical Turk workers who, if they are not in the United States or India, can be paid only in gift tokens — a reinvention of the company store.

Workers for platforms like Freelancer or Elance, despite being firmly told they are not employees, nevertheless find that their clients have acquired exclusive ownership of their intellectual property, even when they have not been paid — a twenty-first century form of wage theft. And tales of online reputations laboriously built up over months or years destroyed in a flash by one spiteful customer’s unchallengeable low star rating.

It appears a new kind of working life is emerging.

It is a life in which who you are and what you can do are displayed to the world in the form of a standardized profile: your skills and the tasks you can perform listed in standard tick-box form, perhaps embellished with some self-promotional text. The strangers with the power to hire you can assess the quality of your work through user ratings that may reflect informed judgement but might equally be an indication of poor taste or a rationale for not paying you.

You don’t know from one week, day, or even hour to the next when or whether you will have work: so keep the smartphone always at hand, ready to hit “accept” at a moment’s notice. You are, in short, permanently logged on.

And since your work is largely carried out online, your every activity is recorded. You are thus continuously generating the data that makes it possible for you to be monitored even more closely, with increasingly precise performance indicators, reducing still further any wiggle room for individual autonomy.

You become part of an atomized workforce, in which individuals are increasingly interchangeable. Their labor is logged: logged in the sense of being chopped up into standardized units; logged in the sense of being connected online, and logged in the sense of being recorded for future analysis. You could call it triply logged.

This might seem overly dystopian. After all, it can be argued, the kinds of work that are now increasingly managed by online platforms have always been precarious. When was taxi-driving or freelance copy-editing ever secure? By what stretch of the imagination was cleaning or running errands ever regarded as a regular job?

One way of looking at the recent exponential growth of online platforms in service delivery is to see it as a formalization of the informal economy, with the transparency of an open market replacing the old word-of-mouth methods of finding work, and the replacement of unrecorded cash-in-hand payments by trackable online payments, opening up at least the possibility for taxes to be collected and fairness to prevail.

A good thing, some might argue. Especially for people who were formerly excluded from such work by gender or ethnicity or disability. Could the techno-utopians be right?

But viewing the online-platform organization of labor as an issue only affecting the informal market is to ignore a larger reality. Several disparate trends that have been slowly building over decades are now converging, accelerating the formation and dissemination of a new labor management model across a range of sectors. It is appearing in both public and private sectors; in manual, clerical, and intellectual jobs; high-skilled and low-skilled; regardless of whether they are covered by permanent employment contracts.

Logged labor is becoming the new norm.

Slowly and insidiously, it has become accepted that you should have an up-to-date resume permanently available for inspection and be ready to pitch yourself anew for each job, promotion, grant, or inclusion in a project team. It is taken for granted that these applications should be made online, requiring you to contort your past experience to fit the standard tick boxes and drop-down menus. Even if you have a contract that specifies a 40-hour week, it is also now normal to expect you to check your email round the clock wherever you happen to be.

Any occasion may be interrupted by a ping on the smartphone, indicating that a summons to a meeting has popped up in your calendar or that a task awaits completion in your inbox. Woe betide you if you forget your username or password when the time comes to act on it: much easier to stay permanently logged in, whatever the security implications.

This is not just a substitution of one kind of communication for another. It is an outward symptom of a major restructuring of work: the manifestation of an underlying pattern whereby tasks are standardized, enabling them to be coordinated and monitored systematically.

Each unit of production is nested into a larger hierarchy of electronically-managed coordination. And each of these units, under pressure to keep costs as low as possible, seeks to minimize them by externalizing as much labor as possible to its users, or the next level down in the hierarchy.

You need a database with everybody’s details in it? Don’t waste money on a data entry clerk. Make all users fill in the online form and enter their own details. You need to be sure that a project will be completed on time? Make all team members log their hours as they go along and introduce penalties for failure to meet targets.

Any given transaction may take only a few minutes or even seconds, but multiplied across a whole economy, having everybody book their own tickets, submit tax returns, upload articles, order groceries, update their profiles, and log their own working hours saves millions of dollars in wages not paid — and adds cumulatively to the cyber-bureaucratic load of unpaid “consumption work” required for everyday survival.

Not only is the cost of this labor externalized to others; these procedures also create an audit trail, allowing each transaction to be tracked, each worker’s performance to be monitored; the basis created for establishing what a “normal” pattern of work should look like for any particular occupational group, which can then be used to set targets for the future.

This model has penetrated many industries and occupations, introducing the paradox of work that is both more formalized — designed to meet standardized performance targets, from volume of content published to sales made, or fine-tuned to function in a complex multinational firm — yet more precarious.

But capitalism requires not just standardization but innovation. And innovation is messy, involving trial and error, sudden bursts of creativity, and false starts.

An increasingly common resolution of this apparent contradiction is, for many capitalist organizations, to put this kind of activity into a sort of black box, hedged around with external controls and involving a minimization of risk.

The trend is towards the development of new products and processes — as well as many aspects of research — to be located in specialist departments or outsourced altogether, with work organized on a project-to-project basis and carried out by temporary teams.

This is a model that has long characterized creative industries, where workers have traditionally come together to produce a particular film, play, or album, and it has spread to video games, software development, and many other applications.

This is where the precariousness comes in. Even if they are ostensibly employees, high-skill development workers are increasingly likely to feel they are only as good as their last project. Each time, they have to prove themselves; putting in extra hours, showing extra dedication, and performing that difficult balancing act of demonstrating that they are a good team player while drawing attention to their individual brilliance — anything to make sure they will be picked for the next team. Life inside the corporation is coming to resemble life outside it ever more closely.

There are even more marked similarities when it comes to low-paid service workers. There is little difference between working on call for a supermarket, warehouse, café, or hamburger chain — waiting for the boss to call you in for work — and watching the smartphone, wondering when the next job from TaskRabbit, Hassle, Handy, or Uber will come in.

One difference between being a freelancer and being an employee is theoretically that the former has more freedom to say no. But in these days of unpaid internships and online theft of intellectual property, for those without rich parents or spouses to lean on, this is increasingly coming to resemble Anatole France’s famous freedom to sleep under bridges and beg in the streets.

Choice only exists when there are genuine alternatives to choose from.

Does this mean, as some have suggested, that all workers are becoming part of a common “precariat” or undifferentiated “multitude”? No. Precariousness is the normal condition of all labor under capitalism — held at bay only by strong organization of workers under favorable circumstances. It is no more a glue that binds workers together in a common class identity than, say, hunger or poverty.

Bringing larger swathes of the workforce into open markets produces more, not less, differentiation. It does not reduce the importance of specific skills and talents. On the contrary, it accentuates it. Employers are able to pinpoint their exact requirements and target potential workers with great precision, setting them in direct competition with each other, often regardless of location.

The workforce is atomized, but workers are generally substitutable for each other only within the constraints of particular skill sets.

Introducing a formalized and atomized market in a situation that has previously been relatively closed brings new winners as well as new losers. What is catastrophic for a freelance graphic designer or a licensed taxi driver in New York or London may well be a life-changing opportunity for an arts graduate in Pakistan or Bolivia or a newly arrived migrant whose only asset is a car. Designating them all as members of a common precarious class does not magically whisk away the real difference in their material interests.

How should political economists understand the new phase of capitalism that is now reaching critical mass? It has several dimensions.

First, it brings new areas of life within the orbit of capital, allowing profits to be made from activities that were previously freely shared. Companies like Airbnb, SnapGoods, Lyft, or the Italian Gnammo, for example, bring human sociality within the orbit of capitalism, allowing a rent to be taken from each act of sharing, even when no employment relationship exists.

Second, whether it involves the online organization of online work (like Elance, Amazon Mechanical Turk, or Upwork) or manual work (like TaskRabbit, Handy, or Zaarly) or markets for petty manufacture (like Etsy) the platform economy extends capitalism’s scope into the informal economy, again taking a hefty rent from each transaction, as well as bringing this labor within the scope of capitalist discipline and time regimes.

Third, it represents an externalization of investment costs. In the past, companies that ran hotel chains or car fleets had to invest in real estate or automobiles, which represented depreciating assets. Now Airbnb, Uber, and the like have persuaded citizens to carry this cost, including the interest on the loans and mortgages they have had to take out to acquire these assets. This ties them ever more tightly into the capitalist system even while they carry the burden of risk.

The fourth impact is perhaps even more insidious because it extends way beyond the scope of the new online platforms into the heartlands of the “old” economy. By establishing a new normative model of what work should be like — logged in all senses — it removes any sense of entitlement to work that is organized differently.

The bodies, minds, and daily lives of the members of this new global labor force are sites of intense contradiction. They are both highly atomized and highly connected with each other. Their tasks are highly specialized, yet they also have more in common with other workers than ever before.

They must be both autonomous and compliant. They must both compete and collaborate. They must be always available but not show signs of fatigue. They must demonstrate past experience and reputation but are still only judged on the last job. There is the potential for new antagonisms as well as new solidarities between them.

What kinds of consciousness are emerging in these conditions? What potential is there for common demands? Should we look back to the normative models of the third quarter of the twentieth century when, at least for the privileged core workforce of the developed world, it was normal to expect a stable job, with health care, paid holidays and a pension? Or has the time come to consider completely different ways of organizing work and welfare?