My argument would be that while the timing is different, programming might not be all the different from other professions in its claims to exclusive mastery based on knowledge of protocols shorn of certain material and practical dimensions. In this regard, is it all that different from architecture? What might need explaining is rather how software intervened in, and transforms, all the professions. Most of them have been redefined as kinds of information-work. In many cases this can lead to deskilling and casualization, on the one hand; and to the circling of the wagons around certain higher-order, but information-based, functions on the other. It is not that programming is an example of “neoliberalism,” so much as that neoliberalism has become a catch-all term for a collection of symptoms of the role of computing in its current form in the production of information as a control layer.

Hence my problem with the ambiguity in formulations such as this: “Software becomes axiomatic. As a first principle, it fastens in place a certain neoliberal logic of cause and effect, based on the erasure of execution and the privileging of programming” (49). What if it is not that software enables neoliberalism, but rather that neoliberalism is just a rather inaccurate way of describing a software-centric mode of production? The invisible machine joins the list of other invisible operators: slaves, women, workers. They don’t need to be all that visible so long as they do what they’re told. They need only to be seen to do what they are supposed to do. Invisibility is the other side of power.13 To the extent that software has power or is power, it isn’t an imaginary fetish.

To what extent is information the missing “complement” to the commodity? There is only one kind of (proto-) information in Marx, and that is the general equivalent—money. The materiality of a thing—let’s say “coats”—its use value, is doubled by its informational quantity, its exchange value, and it is exchanged against the general equivalent, or information as quantity. But notice the missing step. Before one can exchange the thing “coats” for money, one needs the information “coats.” What the general equivalent meets in the market is not the thing but another kind of information—let’s call it the general nonequivalent—a general, shared, agreed-upon kind of information about the qualities of things.24

Putting these sketches together, one might then ask what role computing plays in the rise of a political economy (or a post-political one) in which not only is exchange value dominant over use value, but where use value further recedes behind the general nonequivalent, or information about use value. In such a world, fetishism would be mistaking the body for the information, not the other way around, for it is the information that controls the body.

Thus we want to think bodies matter, lives matter, things matter—when actually they are just props for the accumulation of information and information as accumulation. “Neo”liberal is perhaps too retro a term for a world which does not just set bodies “free” to accumulate property, but sets information “free” from bodies, and makes information property in itself. Perhaps bodies are shaped now by more than one kind of code. Perhaps it is no longer a time in which to use Foucault and Derrida to explain computing, but rather to see them as side effects of the era of computing itself.25