Lessons from the Luddites

In which the spectre of the Luddite software engineer is raised, in an AI-driven future where programming languages become commercially redundant, and therefore take on new cultural significance.

In 1812, Lord Byron dedicated his first speech in the House of Lords to the defence of the machine breakers, whose violent acts against the machines replacing their jobs prefigured large scale trade unionism. We know these machine breakers as Luddites, a movement lead by the mysterious, fictional character of General Ludd, although curiously, Byron doesn’t refer to them as such in his speech. With the topic of post-work in the air at the moment, the Luddite movements are instructive; The movement was comprised of workers finding themselves replaced by machines, left not in a post-work Utopia, but in a state of destitution and starvation. According to Hobsbawm (1952), if Luddites broke machines, it was not through a hatred of technology, but through self-preservation. Indeed, when political economist David Ricardo (1921) raised “the machinery question” he did so signalling a change in his own mind, from a Utopian vision where the landlord, capitalist, and labourer all benefit from mechanisation, to one where reduction in gross revenue hits the labourer alone. Against the backdrop of present-day ‘disruptive technology’, the machinery question is as relevant as ever.

The Leader of the Luddites
The Leader of the Luddites

A few years after his speech, Byron went on to father Ada Lovelace, the much celebrated prototypical software engineer. Famously, Ada Lovelace cooperated with Charles Babbage on his Analytical Engine; Lovelace exploring abstract notions of computation at a time when Luddites were fighting against their own replacement by machines. This gives us a helpful narrative link between mill workers of the industrial revolution, and software engineers of the information revolution. That said, Byron’s wayward behaviour took him away from his family, and he deserves no credit for Ada’s upbringing. Ada was instead influenced by her mother Annabella Byron, the anti-slavery and women’s rights campaigner, who encouraged Ada into mathematics.

Today, general purpose computing is becoming as ubiquitous as woven fabric, and is maintained and developed by a global industry of software engineers. While the textile industry developed out of worldwide practices over millennia, deeply embedded in culture, the software industry has developed over a single lifetime, the practice of software engineering literally constructed as a military operation. Nonetheless, the similarity between millworkers and programmers is stark if we consider weaving itself as a technology. Here I am not talking about inventions of the industrial age, but the fundamental, structural crossing of warp and weft, with its extremely complex, generative properties to which we have become largely blind since replacing human weavers with powerlooms and Jacquard devices. As Ellen Harlizius-Klück argues, weaving has been a digital art from the very beginning.

Software engineers are now threatened under strikingly similar circumstances, thanks to breakthroughs in Artificial Intelligence (AI) and “Deep Learning” methods, taking advantage of the processing power of industrial-scale server farms. Jen-Hsun Hu, chief executive of NVIDIA who make some of the chips used in these servers is quoted as saying that now, “Instead of people writing software, we have data writing software”. Too often we think of Luddites as those who are against technology, but this is a profound misunderstanding. Luddites were skilled craftspeople working with technology advanced over thousands of years, who only objected once they were replaced by technology. Deep learning may well not be able to do everything that human software engineers can do, or to the same degree of quality, but this was precisely the situation in the industrial revolution. Machines cannot make the same woven structures as hands, to the same quality, or even at the same speed at first, but the Jacquard mechanism replaced human drawboys anyway.

As a thought experiment then, let’s imagine a future where entire industries of computer programmers are replaced by AI. These programmers would either have to upskill to work in Deep Learning, find something else to do, or form a Luddite movement to disrupt Deep Learning algorithms. The latter case might even seem plausible when we recognise the similarities between the Luddite movement and Anonymous, both outwardly disruptive, lacking central organisation, and lead by an avatar: General Ludd in the case of the Luddites, and Guy Fawkes in the case of Anonymous.

Let’s not dwell on Anonymous though. Instead try to imagine a Utopia in which current experiments in Universal Basic Income are proved effective, and software engineers are able to find gainful activity without the threat of destitution. The question we are left with then is not what to do with all the software engineers, but what to do with all the software? With the arrival of machine weaving and knitting, many craftspeople continued hand weaving and handknitting in their homes and in social clubs for pleasure rather than out of necessity. This was hardly a surprise, as people have always made fabric, and indeed in many parts of the world handweaving has remained the dominant form of fabric making. Through much of the history of general purpose computing however, any cultural context for computer programming has been a distant second to its industrial and military contexts. There has of course been a hackerly counter-culture from the beginning of modern-day computing, but consider that the celebrated early hackers in MIT were funded by the military while Vietnam flared, and the renowned early Cybernetic Serendipity exhibition of electronic art included presentations by General Motors and Boeing, showing no evidence of an undercurrent of political dissent. Nonetheless, I think a Utopian view of the future is possible, but only once Deep Learning renders the craft of programming languages useless for such military and corporate interests.

Looking forward, I see great possibilities. All the young people now learning how to write code for industry may find that the industry has disappeared by the time they graduate, and that their programming skills give no insight into the workings of Deep Learning networks. So, it seems that the scene is set for programming to be untethered from necessity. The activity of programming, free from a military-industrial imperative, may become dedicated almost entirely to cultural activities such as music-making and sculpture, augmenting human abilities to bring understanding to our own data, breathing computational pattern into our lives. Programming languages could slowly become closer to natural languages, simply by developing through use while embedded in culture. Perhaps the growing practice of Live Coding, where software artists have been developing computer languages for creative coding, live interaction and music-making over the past two decades, are a precursor to this. My hope is that we will begin to think of code and data in the same way as we do of knitting patterns and weaving block designs, because from my perspective, they are one and the same, all formal languages, with their structures intricately and literally woven into our everyday lives.

Joanne Armitage Live Coding
Joanne live coding at access space

So in order for human cultures to fully embrace the networks and data of the information revolution, perhaps we should take lessons from the Luddites. Because they were not just agents of disruption, but also agents against disruption, not campaigning against technology, but for technology as a positive cultural force.

This article was written by Alex while sound artist in residence in the Open Data Institute, London, as part of the Sound and Music embedded programme.


Gerald Raunig. Dividuum. Book Review

Gerald Raunig’s book on the concept and the genealogy of “dividuum” is a learned tour de force about what is supposed to be replacing the individual, if the latter is going to “break down” – as Raunig suggests. The dividuum is presented as a concept with roots in Epicurean and Platonic philosophy that has been most controversially disputed in medieval philosophy and achieves new relevance in our days, i.e. the era of machinic capitalism.

The word dividuum appears for the first time in a comedy by Plautus from the year 211 BC . “Dividuom facere” is Ancient Latin for making something divisible that has not been able to be divided before. This is at the core of dividuality that is presented here as a counter-concept to individuality.

Raunig introduces the notion of dividuality by presenting a section of the comedy “Rudens” by Plautus. The free citizen Daemones is faced with a problem: His slave Gripus pulls a treasure from the sea that belongs to the slave-holder Labrax and demands his share. Technically the treasure belongs to the slave-holder, and not to the slave. Yet, in terms of pragmatic reasoning there would have been no treasure if the slave hadn’t pulled it from the sea and declared it as found to his master. Daemones comes up with a solution that he calls “dividuom facere”. He assigns half of the money to Labrax, who is technically the owner, if he sets free Ampelisca, a female slave whom Gripus loves.  The other half of the treasure is assigned to Gripus for buying himself out of slavery. This will allow Gripus to marry Ampelisca and it will also benefit the patriarch’s wealth. Raunig points out that Daemones transfers formerly indivisible rights and statuses – that of the finder and that of the owner – into a multiple assignment of findership and ownership.

We will later in the book see how Raunig sees this magic trick being reproduced in contemporary society, when we turn or are turned into producer/consumers, photographers/photographed, players/instruments, surveyors/surveyed under the guiding imperative of “free/ voluntary obedience [freiwilliger Gehorsam]”.

The reader need not be scared by text sections in Ancient Greek, Ancient Latin, Modern Latin or Middle High German, as Raunig provides with translations. The translator Aileen Derieg must be praised for having undertaken this difficult task once more for the English language version of the book. It can however be questioned whether issues of proper or improper translation of Greek terms into Latin in Cicero’s writings in the first century BC should be discussed in the context of this book. This seems to be a matter for an old philology publication, rather than for a Semiotext(e) issue.

It is, however, interesting to see how Raunig’s genealogy of the dividuum portraits scholastic quarrels as philosophic background for the development of the dividuum. With the trinity of God, Son of God and the Holy Spirit the Christian dogma faced a problem of explaining how a trinity and a simultaneous unity of God could be implemented in a plausible and consistent fashion. Attempts to speak of the three personae (masks) of God or to propose that father, son and spirit were just three modes (modalism) of the one god led to accusations of heresy. Gilbert de Poitiers’ “rationes” were methodological constructions that claimed at least two different approaches to reasoning: One in the realm of the godly and the other in the creaturely realm. Unity and plurality could in this way be detected in the former and the latter realm respectively without contradicting each other. Gilbert arrives at a point where he concludes that “diversity and conformity are not to be seen as mutually exclusive, but rather as mutually conditional.” Dissimilarity is a property of the individual that forms a whole in space and time. The dividual, however, is characterized by similarity, or in Gilbert de Poitiers’ words: co-formity. “Co-formity means that parts that share their form with others assemblage along the dividual line. Dividuality emerges as assemblage of co-formity to form-multiplicity. The dividual type of singularity runs through various single things according to their similar properties. … This co-formity, multi-formity constitutes the dividual parts as unum dividuum.”

In contemporary society the one that can be divided, unum dividuum, can be found in “samples, data, markets, or ‘banks’”. Developing his argument in the footsteps of Gilles Deleuze, Gerald Raunig critically analyses Quantified Self, Life-Logging, purchases from Amazon, film recommendations from Netflix, Big Data and Facebook. “Facebook needs the self-division of individual users just as intelligence agencies continue to retain individual identities. Big Data on the other hand, is less interested in individuals and just as little interested in a totalization of data, but all the more so in data sets that are freely floating and as detailed as possible, which it can individually traverse.” In the beginning “Google & Co. tailor search results individually (geographical position, language, individual search history, personal profile, etc.” The qualitative leap from individual data to dividual data happens when the ”massive transition to machinic recommendation [happens], which save the customer the trouble of engaging in a relatively undetermined search process.” With this new form of searching the target of my search request is generated by my search history and by an opaque layer of preformatted selection criteria. The formerly individual subject of the search process turns into a dividuum that is asking for data and providing data, that is replying to the subject’s own questions and becomes interrogator and responder in one. This is where Gilbert de Poitiers’ unum dividuum reappers in the disguise of contemporary techno-rationality.

Raunig succeeds in creating a bridge from ancient and medieval texts to contemporary technologies and application areas. He discusses Arjun Appadurai’s theory of the “predatory dividualism”, speculates about Lev Sergeyevich Termen’s dividual machine from the 1920s that created a player that is played by technology, and theorizes on “dividual law” as observed in Hugo Chávez’ presidency or the Ecuadorian “Constituyente” from 2007. The reviewer mentions these few selected fields of Raunig’s research to show that his book is not far from an attempt to explain the whole world seamlessly from 210 BC to now, and one can’t help feeling overwhelmed by the text.

For some readers Raunig’s manneristic style of writing might be over the top, when he calls the paragraphs of his book “ritornelli”, when he introduces a chapter in the middle of his arguments that just lists people whom he wants to thank, or when he poetically goes beyond comprehensiveness and engages with grammatical extravaganzas in sentences like this one: “[: Crossing right through space, through time, through becoming and past. Eternal queer return. An eighty-thousand or roughly 368-old being traversing multiple individuals.”

It is obvious that Gerald Raunig owes the authors of the Anti-Œdipe stylistic inspirations, theoretical background and a terminological basis, when he talks of the molecular revolutions, desiring machines, the dividuum, deterritorialization, machinic subservience and lines of flight. He manages to contextualise a concept that has historic weight and current relevance within a wide field of cultural phenomena, philosophical positions, technological innovation and political problems. A great book.

Machinic Capitalism and Molecular Revolution
By Gerald Raunig
Translated by Aileen Derieg
Semiotext(e) / MIT Press

Guido Segni’s A quiet desert failure

The new project by Guido Segni is so monumental in scope and so multitudinous in its implications that it can be a bit slippery to get a handle on it in a meaningful way. A quiet desert failure is one of those ideas that is deceptively simple on the surface but look closer and you quickly find yourself falling down a rabbit-hole of tangential thoughts, references, and connections. Segni summarises the project as an “ongoing algorithmic performance” in which a custom bot programmed by the artist “traverses the datascape of Google Maps in order to fill a Tumblr blog and its datacenters with a remapped representation of the whole Sahara Desert, one post at a time, every 30 minutes.”1

Opening the Tumblr page that forms the core component of A quiet desert failure it is hard not to get lost in the visual romanticism of it. The page is a patchwork of soft beiges, mauves, creams, and threads of pale terracotta that look like arteries or bronchia. At least this morning it was. Since the bot posts every 30 minutes around the clock, the page on other days is dominated by yellows, reds, myriad grey tones. Every now and then the eye is captured by tiny remnants of human intervention; something that looks like a road, or a small settlement; a lone, white building being bleached by the sun. The distance of the satellite, and thus our vicarious view, from the actual terrain (not to mention the climate, people, politics, and more) renders everything safely, sensuously fuzzy; in a word, beautiful. Perhaps dangerously so.

As is the nature of social media platforms that prescribe and mediate our experience of the content we access through them, actually following the A quiet desert failure Tumblr account and encountering each post individually through the template of the Tumblr dashboard provides a totally different layer to the work. On the one hand this mode allows the occasional stunningly perfect compositions to come to the fore – see image below – some of these individual ‘frames’ feel almost too perfect to have been lifted at random by an aesthetically indifferent bot. Of course with the sheer scope of visual information being scoured, packaged, and disseminated here there are bound to be some that hit the aesthetic jackpot. Viewed individually, some of these gorgeous images feel like the next generation of automated-process artworks – a link to the automatic drawing machines of, say, Jean Tinguely. Although one could also construct a lineage back to Duchamp’s readymades.

Segni encourages us to invest our aesthetic sensibility in the work. On his personal website, the artist has installed on his homepage a version of A quiet desert failure that features a series of animated digital scribbles overlaid over a screenshot of the desert images the bot trawls for. Then there is the page which combines floating, overlapping, translucent Google Maps captures with an eery, alternately bass-heavy then shrill, atmospheric soundtrack by Fabio Angeli and Lorenzo Del Grande. The attention to detail is noteworthy here; from the automatically transforming URL in the browser bar to the hat tip to themes around “big data” in the real time updating of the number of bytes of data that have been dispersed through the project, Segni pushes the limits of the digital medium, bending and subverting the standardised platforms at every turn.

But this is not art about an aesthetic. A quiet desert failure did begin after the term New Aesthetic came to prominence in 2012, and the visual components of the work do – at least superficially – fit into that genre, or ideology. Thankfully, however, this project goes much further than just reflecting on the aesthetic influence of “modern network culture”2 and rehashing the problematically anthropocentric humanism of questions about the way machines ‘see’. Segni’s monumental work takes us to the heart of some of the most critical issues facing our increasingly networked society and the cultural impact of digitalisation.

The Sahara Desert is the largest non-polar desert in the world covering nearly 5000 km across northern Africa from the Atlantic ocean in the west to the Red Sea in the east, and ranging from the Mediterranean Sea in the north almost 2000 km south towards central Africa. The notoriously inhospitable climate conditions combine with political unrest, poverty, and post-colonial power struggles across the dozen or so countries across the Sahara Desert to make it surely one of the most difficult areas for foreigners to traverse. And yet, through the ‘wonders’ of network technologies, global internet corporations, server farms, and satellites, we can have a level of access to even the most problematic, war-torn, and infrastructure-poor parts of the planet that would have been unimaginable just a few decades ago.

A quiet desert failure, through the sheer scope of the piece, which will take – at a rate of one image posted every 30 minutes – 50 years to complete, draws attention to the vast amounts of data that are being created and stored through networked technologies. From there, it’s only a short step to wondering about the amount of material, infrastructure, and machinery required to maintain – and, indeed, expand – such data hoarding. Earlier this month a collaboration between private companies, NASA, and the International Space Station was announced that plans to launch around 150 new satellites into space in order to provide daily updating global earth images from space3. The California-based company leading the project, Planet Labs, forecasts uses as varied as farmers tracking crops to international aid agencies planning emergency responses after natural disasters. While it is encouraging that Planet Labs publishes a code of ethics4 on their website laying out their concerns regarding privacy, space debris, and sustainability, there is precious little detail available and governments are, it seems, hopelessly out of date in terms of regulating, monitoring, or otherwise ensuring that private organisations with such enormous access to potentially sensitive information are acting in a manner that is in the public interest.

The choice of the Sahara Desert is significant. The artist, in fact, calls an understanding of the reasons behind this choice “key to interpret[ing] the work”. Desertification – the process by which an area becomes a desert – involves the rapid depletion of plant life and soil erosion, usually caused by a combination of drought and overexploitation of vegetation by humans.5 A quiet desert failure suggests “a kind of desertification taking place in a Tumblr archive and [across] the Internet.”6 For Segni, Tumblr, more even than Instagram or any of the other digitally fenced user generated content reichs colonising whatever is left of the ‘free internet’, is symbolic of the danger facing today’s Internet – “with it’s tons of posts, images, and video shared across its highways and doomed to oblivion. Remember Geocities?”7

From this point of view, the project takes on a rather melancholic aspect. A half-decade-long, stately and beautiful funeral march. An achingly slow last salute to a state of the internet that doesn’t yet know it is walking dead; that goes for the technology, the posts that will be lost, the interior lives of teenagers, artists, nerds, people who would claim that “my Tumblr is what the inside of my head looks like”8 – a whole social structure backed by a particular digital architecture, power structure, and socio-political agenda.

                                            a quiet desrt failre tumblr.com/archive

The performative aspect of A quiet desert failure lies in the expectation of its inherent breakdown and decay. Over the 50 year duration of the performance – not a randomly selected timeframe, but determined by Tumblr’s policy regulating how many posts a user can make in a day – it is likely that one or more of the technological building blocks upon which the project rests will be retired. In this way we see that the performance is multi-layered; not just the algorithm, but also the programming of the algorithm, and not just that but the programming of all the algorithms across all the various platforms and net-based services incorporated, and not just those but also all the users, and how they use the services available to them (or don’t), and how all of the above interact with new services yet to be created, and future users, and how they perform online, and basically all of the whole web of interconnections between human and non-human “actants” (as defined by Actor- network theory) that come together to make up the system of network, digital, and telecommunications technologies as we know them.

Perhaps the best piece I know that explains this performativity in technology is the two-minute video New Zealand-based artist Luke Munn made for my Net Work Compendium – a curated collection of works documenting the breadth of networked performance practices. The piece is a recording of code that displays the following text, one word at a time, each word visible for exactly one second: “This is a performance. One word per second. Perfectly timed, perfectly executed. All day, every day. One line after another. Command upon endless command. Each statement tirelessly completed. Zero one, zero one. Slave to the master. Such was the promise. But exhaustion is inevitable. This memory fills up. Fragmented and leaking. This processor slows down. Each cycle steals lifecycle. This word milliseconds late. That loop fractionally delayed. Things get lost, corrupted. Objects become jagged, frozen. The CPU is oblivious to all this. Locked away, hermetically sealed, completely focused. This performance is always perfect.”

Guido Segni’s A quiet desert failure is, contrary to its rather bombastic scale, a finely attuned and sensitively implemented work about technology and our relationship to it, obsolescence (planned and otherwise), and the fragility of culture (notice I do not write “digital” culture) during this phase of rapid digitalisation. The work has been released as part of The Wrong – New Digital Art Biennale, in an online pavilion curated by Filippo Lorenzin and Kamilia Kard, inexactitudeinscience.com.


  1. Source: http://guidosegni.com/work/a-quiet-desert-failure/
  2. Sterling, Bruce. “An Essay on the New Aesthetic,” Beyond The Beyond, Wired (2012)
  3. Baddour, Dylan. “Massive satellite web will produce daily images of the entire Earth,” Chron, http://www.chron.com/news/nation-world/space/article/NASA-Planet-Labs-satellites-ISS-discovery- space-6625269.php
  4. See https://www.planet.com/ethics/
  5. Source: dictionary.com – http://dictionary.reference.com/browse/desertification?s=t
  6. Segni, Guido, via email to Gretta Louw, 16/11/2015.
  7. Ibid.
  8. For example: http://mizzizbananas.tumblr.com