Who Makes The Machines That Make Machines?

Content Warning: Another essay couched in a book review…

Michael Graaf, Feb 2023


The book is Exodus: General Idea of the Revolution in the XXI Century | Kevin Carson. One of
those books written as if people still read books. Well of course, some of us still try to, but it’ll
never be the way it was before tl:dr became a thing.

The essay aspect presents a contemporary context for reading the book: just as the North Atlantic
empire found it necessary to destroy the NordStream pipelines to prevent symbiosis between
Europe and Russia, it is also willing to destroy Taiwan’s chip foundries to prevent them from falling
into “communist” hands. Also, a glimpse of why this could become self-defeating, just as the attack
on Russia’s economy was.


The book starts off with a review of the idea of the (anti-capitalist) revolution in Marxist and anarchist
traditions, i.e. Euro-American 19th and 20th century writings. This review (like the book as a whole) is analytically Marxist (i.e. applying historical materialism in describing writers, writings, and the movements they were embedded in) but politically anarchist (as you’d expect in a text from the Centre for a Stateless Society); its default treatment of the state is as something extraneous.

To tl;dr the review: the age of steam & coal fostered not only mass-industrial production (with accompanying nationalist politics), but mass-industrial resistance – such as trade unionism and electoral “labour” parties. The shift to electricity and purer hydrocarbons didn’t disrupt these. Luddite and anarchist activists were doomed to marginality in these circumstances.

Part 2 of the book examines the emergence and effects of digital technologies. The possibilities
enabled by many-to-many communication are already well understood, but another crucial concept
is that of stigmergy: the prompting of tasks by the effects of previous tasks. A grand project like the
development of an operating system or the care and feeding of a population, needn’t always or
completely depend on a shared overall plan or vision if participants are able to assess a local context
and respond with local action which happens to contribute to the grand project.

Stigmergy is not “collectivist” in the traditional sense, as it was understood
in the days when a common effort on any significant scale required a large or-
ganization to represent the collective, and the administrative coordination of
individual efforts through a hierarchy. But it is the ultimate realization of col-
lectivism, in that it removes the transaction costs involved in concerted action
by many individuals. (Exodus, p74)

Some critique:

Carson rightly highlights the strategic importance of the rapid decline in costs not only of ICTs but
also energy systems. However he tends to assume that this process, if it hasn’t already made these
things accessible to (almost) everybody, will soon do so. This shows his privileged position within
the world’s population. If digital tools with the humble (yet, for the time, revolutionary) abilities of
the early PCs were still made, they would by now be affordable by almost everybody. However,
they are no longer made. Consider the phenomenon wryly described by the saying “Software grows
slower, faster than hardware grows faster.” Both processes result from the commercial drive to
capture middle-to-high spenders, leaving the lower half either stranded or else forever struggling to
catch up, using hand-me-down equipment which all too soon is no longer supported or repairable.
Although free and open-source software projects and communities mitigate this, they themselves
also have to play catch-up. Nor is energy equipment immune: for example, as time goes by more
and more batteries come with integrated microprocessors, displays, apps etc.

Another aspect of global privilege is the taking for granted of the digital components of the new
tools. These have long been (when not already assembled into consumer goods) easily ordered from
companies that mass-produce them. All that is now starkly questionable with the imposition of
sanctions on China. The geopolitics of chip production is now front and centre as never before.


Carson goes on to develop the central thesis of the book – that new, decentralised technologies not
only enable but even foster the growth “in the interstices” (gaps or cracks) of capitalism, of its
successor. Not that he suggests its inevitability – he does an exhaustive review of new
organisational and political cultures needed to usher it in.

Although suggesting that the members/supporters of the new system won’t need to seize control of
the old one, Carson concedes that there will probably be some kicks of the dying horse which may
require self-defence by the post-capitalists, along with their process of commoning (returning once-
hoarded resources to the public realm). And at one point he sounds almost like Thomas Sankara:

“ capitalism becomes increasingly dependent on credit expansion, investment bubbles
and the FIRE economy for maintaining aggregate demand, it becomes to that extent
vulnerable to other mass actions like debt strikes by the population of the imperial core, and
coordinated national debt defaults by debtor nations of the Global South.”

More critique:

What about the half of humanity currently and for the foreseeable future without access to tech-
utopian tools and visions? Certainly, something like the more resource-efficient economy Carson
envisages would be necessary if all humanity is to thrive. The question is whether it’s sufficient.
Can we really get by without forcefully ending the artificial shortages imposed by capitalism?
Could it happen that the trajectory he describes occurs in the “core” imperial countries, leaving the
rest of us at the mercy of warlords? Let me illustrate this fear using terms used in the text, quoting

“The problem is not to destroy that society but to stop creating it. Capitalism exists today
not because we created [it] two hundred years ago or a hundred years ago, but because we
create it today. If we do not create it tomorrow, it will not exist.”

That is spoken as a citizen of the core. Out here in the neocolonies, capitalism was never made but
rather arrived, and is still actively destroying the remains of the precapitalist economy (rural
subsistence). Large numbers of people are perched precariously on the narrow ledges of social
grants, without a safety net below. Or, to use an image from the news, are in flimsy boats
approaching the cliffs of Fortress Europe, and the rescue vessels are already overloaded.

Carson’s book acknowledges this reality, but copes with it by glorifying the Zapatistas, the MST
etc. which are largely rearguard actions against the advance of capital in the periphery.

A way forward?

To indulge in some whimsical speculation – perhaps we need to revisit the “appropriate technology”
paradigm of the 1970s & 80s. Indeed, Carson hints at this in reviewing how things have changed
since the publication of Small is Beautiful, and how things could be as a result. He conjures a
world of makers sharing designs globally while producing locally needed items locally, along with
platform co-operatives enabling a true sharing economy where “deathstars” like Uber et al can’t

But what about people still in the queue to enter the digital age? This vignette may help develop that
thought: the Raspberry Pi project from its inception explicitly harked back to the BBC Micro
computer project; both aimed to make computing accessible to young people. Yet as it matched,
then surpassed the success of its predecessor, Raspberry Pi turned to a ‘lower-tech’ project: the
RP2040. This custom microprocessor actually has the capacity of the original BBC Micro (within a
few days of the RP2040’s release, somebody wrote a BBC Micro emulator to run on it), but is
vastly more affordable (USD 1!) even than the R Pi, never mind the Micro, and also has an open
architecture – something that was not possible when the Micro and R Pi were cobbled together with
off-the-shelf components. It also uses a fraction of the energy and materials to build and
operate1. Imagine, for example, a low-cost, low-energy computer lab or telecentre using RP2040-
based thin clients booting from an instructor/supervisor’s laptop, the latter being safely removed
when not in use, leaving not much worth stealing, because the components have little resale value;
they only work as a whole, not as separate units. Imagine it deployed in tens of thousands of schools which currently lack libraries or internet facilities: the Model T of local area networks (or dare we say, the AK-47).

The other speculative thread here is the RISC-V (“risk-5”) architecture. Just as a global trade war is
hotting up at the top end of the silicon market, global collaboration is surging lower down. While
not immediately able to enter the arena of high-performance devices, RISC-V offers to stimulate
competition and lower entry barriers in the long run, due to its modular character and public

RISC-V could do to the world of hardware what Linux did to the world of software.

Its history of international co-operation also provides a foundation of goodwill that could survive the corporate bandwagon-jumping that has already started. Having been a founding partner, China as victim of the above-mentioned trade war, is redoubling its contribution to the project – benefiting the entire
world. Just as an oil embargo worked against Venezuela but backfired when applied to Russia (it
hastened the demise of the petrodollar), a chip embargo which would cripple Korea if applied there,
may end up strengthening not only China but the global open-licensed silicon movement.

Postscript: Just over a year later, everything RISC-V is going nicely. As long as we avoid nuclear war there’s hope.

1Furthermore, the main product incorporating the RP2040, namely the Pico, is now assembled in Kenya (other R Pi products are assembled in Wales).