Software is a Service

There’s a theory of software’s place in political economy that I’ve been considering for a little while now. It tries to make sense of what software is doing in and for the economy today. The focus is on the American context.

My basic premise is this: software works in a capitalist economy because it’s fundamentally a commodity. The code or “IP” isn’t the commodity, though; it’s the experience that code produces for customers and users. The business that owns the software use that software to provide its service.

The story to get to this point is, from one perspective, quite simple. Computers compute. That’s an action and the whole computing industry is about doing that with various technical objects for a fee. Software is the means by which computers are programmed to compute various functions.

From another perspective, the story is anything but simple. Software as a concept was born in the 20th century. Numerous historians and cultural scholars have traced it from cybernetics and the development of information theory. Others go back to places like the census and the need to produce statistical models of populations. Still others Babbage and Lovelace. Yet others Leibniz. We’ll start with cybernetics. At first it’s academic and “pure” mathematics. Weiner and Shannon write reams and reams of formulae, and those are the theoretic underpinnings of new sectors in the military and private industry. Somehow that makes its way into “applied” territory and you start getting early computer programmers. There are still lots of formulae, there is still lots of computing. But there’s something else going on.

The economy is undergoing a massive shift. Manufacturing, the mass production of saleable physical goods, is becoming more and more expensive. Organized and unorganized labor resists the imperatives of capital at home, and anti-colonial struggles threaten supply lines abroad. These disruptions make it hard to feel confident in making a buck. So capital starts the process of off-shoring and making in-roads through the Iron Curtain. Slowly but surely the ratio of goods to services produced in America shifts.

The economic geography of the country changes. Industrial powerhouses are wrecked and become post-industrial shells. A certain portion of the population (the Boomers) has it really good, getting it coming and going. They inherit a strong manufacturing economy and based on that security start craving new experiences. That cohort spends their money to create those experiences for themselves and their progeny. Those progeny then have to keep it up because the Boomers still control the vast majority of the wealth. Gen X and Y (Millennials) live through increasing austerity as more and more capital goes towards creating remarkable services, which increasingly rely on computers and software. Millennials and Gen Z get there just in time for the internet to become a private utility and their interactions with various websites and apps both engages them in new experiences as users and creates the data necessary for these new businesses to run it through analytic software that informs them how to make more successful services (in the aggregate). The indeterminacy of “AI” makes this all even more chaotic and produces even more data and it all just spirals.

How do these services work? Software is composed of programs that tell computers how to route electricity through hardware in ways that send specific amounts of energy to specific places in order to produce certain effects. In the old times that effect could be moving a cursor such that it wrote on paper; now “printing” includes other operations like manipulating lasers. Another way that software provides a service is lighting up a screen in certain patterns. These patterns are constructed from the light projected through pixelated glass. As the light projects, each pixel “informs” the observer and their body turns it into a coherent image. Then the pixels show lights of different colors and shades the observer must make sense anew. What this means is that screens are constantly bombarding people with information and making them react and react and react. Looking at a screen and making it make sense takes effort, and each time it’s a genuinely new experience designed to seem like a commodity.

From the business side, most software developers don’t really need to work at that level of detail any more. They have software bundled in software bundled in software. Just tons of leaky layers of nested programs and abstractions. Especially in the era of cloud compute and Web 2.0, the businesses increasingly try to spend their R&D money as Operational Expenditures (OpEx) and not Capital Expenditures (CapEx). The cloud lets them buy the compute they need as a service (a rental) rather than buy and run the hardware themselves. This is generally preferable because it grants them greater liquidity and optionality as to how they’re deploy that resource.

Furthermore, software companies want lots of users creating lots of data that they can process and offer as information to their customers. To get that they need to provide a commodified experience to their users and their customers; those may be one and the same, as in the case of enterprise software. The role of a software developer is increasingly “shifting left” and thus the job is to manage the flexibility and interoperation of the changing fleets of computers and programs to create a commodified experience. In other words, these software developers are service providers and work in the service sector.

Where will software and the service economy of America go from here? Who knows. Just throwing this out there.

Previous
Previous

The Economy of Thought

Next
Next

The Differing Roles of Mathematics in Engineering