As the object of study rather than the subject of communication, the so-called Middle East has long been a locus for advanced technologies of mapping. In the field of aerial vision, these technologies historically employed cartographic and photographic methods. The legacy of cadastral, photographic and photogrammetric devices continues to impact how people and regions are quantified, nowhere more so than when we consider the all-encompassing, calculative gaze of autonomous systems of surveillance. Perpetuated and maintained by Artificial Intelligence (AI), these remotely powered technologies herald an evolving era of mapping that is increasingly implemented <\/span>through<\/span><\/i> the operative logic of algorithms.\u00a0<\/span><\/p>\n Algorithmically powered models of data extraction and image processing have, furthermore, incrementally refined neo-colonial objectives: while colonization, through cartographic and other less subtle means, was concerned with wealth and labour extraction, neo-colonization, while still pursuing such objectives, is increasingly preoccupied with data extraction and automated models of predictive analysis. Involving as it does the algorithmic processing of data to power machine learning and computer vision, the functioning of these predictive models is indelibly bound up with, if not encoded by, the martial ambition to calculate, or forecast, events that are <\/span>yet to happen<\/span><\/i>.\u00a0<\/span><\/p>\n As a calculated approach to <\/span>re<\/span><\/i>presenting people, communities and topographies, the extraction and application of data is directly related to computational projection: \u2018The emphasis on number and the instrumentality of knowledge has a strong association with cartography as mapping assigns a position to all places and objects. <\/span>That position can be expressed numerically<\/span><\/i>.\u2019<\/span> If a place or object can be expressed numerically, it bestows a privileged command on to the colonial I\/eye of the cartographer. This positionality can be readily deployed to manage \u2013 regulate, govern and occupy \u2013 and contain the present and, inevitably, the future.<\/span><\/p>\n These panoptic and projective ambitions, initially embodied in the I\/eye of the singular embodied figure of the cartographer, nevertheless need to be automated if they are to remain strategically viable. To impose a perpetual command entails the development of increasingly capacious models of programmed perception.<\/span> \u00a0Established to support the aspirations of colonialism and the imperatives of neo-colonial military-industrial complexes, contemporary mapping technologies \u2013 effected through the mechanized affordances of AI \u2013 extract and quantify data in order to project it back onto a given environment. The overarching effect of these practices of computational projection is the de facto expansion of the all-seeing neo-colonial gaze <\/span>into the future<\/span><\/i>.<\/span><\/p>\n The evolution of remote, disembodied technologies of perpetual surveillance, drawing as they did upon the historical logic and logistics of colonial cartographic methods, also necessitated the transference of sight \u2013 the ocular-centric event of seeing and perception \u2013 to the realm of the machinic. The extractive coercions and projective compulsions of colonial power not only saw the entrustment of sight to machinic models of perception but also summoned forth the inevitable automation of image production. It is within the context of Harun Farocki\u2019s \u2018operational images\u2019, by way of Vil\u00e9m Flusser\u2019s theorization of \u2018technical images\u2019, that we can link the colonial ambition to automate sight with the role performed by AI in the extractive pursuits of neo-colonialism.<\/span><\/p>\n Harun Farocki, Eye Machine III, \u00a9 Harun Farocki, 2003. Image via Springerin<\/p>\n<\/div>\n Based as they are on alignments <\/span>within<\/span><\/i> processes of automation, mining, quantifying and archiving, \u2018technical images\u2019 and \u2018operational images\u2019 foreshadow methods of data retrieval, storage and targeting, which are now associated with the algorithmic apparatuses that power Unmanned Aerial Vehicles (UAV) and Lethal Autonomous Weapons (LAW). When we consider the relationship between \u2018technical images\u2019 and \u2018operational images\u2019, in the context of the devolution of ocular-centric models of vision and automated image processing, we can also more readily recognize how the deployment of AI in UAVs and LAWs propagates an apparatus of dominion that both <\/span>contains<\/span><\/i> and <\/span>suspends<\/span><\/i> the future, especially those futures that do not serve the imperatives of neo-colonization.<\/span><\/p>\n Understood as a system that demonstrates non-human agency, for Flusser an apparatus essentially \u2018simulates\u2019 thought and, via computational processes, enables models of automated image production to emerge. <\/span>The \u2018technical image\u2019 is, accordingly, \u2018an image produced by apparatuses\u2019 \u2013 the outcome of sequenced, recursive computations rather than human-centric activities. <\/span>\u2018In this way,\u2019 Flusser proposed, \u2018the original terms <\/span>human<\/span><\/i> and <\/span>apparatus<\/span><\/i> are reversed, and human beings operate as a function of the apparatus.\u2019<\/span> It is the sense of an autonomous machinic functioning behind image production that informs Harun Farocki\u2019s seminal account of the \u2018operational image\u2019.\u00a0<\/span><\/p>\n Void of aesthetic context, \u2018operational images\u2019 are part of a machine-based operative logic and do not, in Farocki\u2019s words, \u2018portray a process but are themselves part of a process.\u2019 <\/span> Indelibly defined by the operation in question, rather than any referential logic, these images are not propagandistic (they do not try to convince), nor are they levelled towards the ocular-centric realm of human sight (they are not, for example, interested in directing our attention). Inasmuch as they exist as abstract binary code rather than pictograms, they are likewise not imagistic \u2013 in fact, they are not even images: \u2018A computer can process pictures, but it needs no pictures to verify or falsify what it reads in the images it processes.\u2019<\/span> Reduced to numeric code, \u2018operational images\u2019, in the form of sequenced code or vectors, remain foundational to the development of contemporary models of recursive machine learning and computer vision.<\/span><\/p>\n In the final part of Farocki\u2019s <\/span>Eye\/Machine<\/span><\/i> trilogy (2001\u20132003) there is a conspicuous focus on the non-allegorical, recursively relayed image \u2013 the \u2018operational image\u2019 \u2013 and its role in supporting contemporary models of aerial targeting.<\/span>\u00a0 <\/span>In direct reference to the first Gulf War in 1991, and the subsequent invasions of Afghanistan and Iraq in 2001 and 2003, Farocki observed that a \u2018new policy on images\u2019 had ushered in a paradigm of opaque and largely unaccountable methods of image production that would inexorably inform the future of \u2018electronic warfare\u2019.<\/span><\/p>\n Harun Farocki, Eye Machine III, \u00a9 Harun Farocki, 2003. Image via Springerin<\/p>\n<\/div>\n The <\/span>novelty<\/span><\/i> of the operational images in use in 2003 in Iraq, it has been argued, \u2018lies in the fact that they were not originally intended to be seen by humans but rather were supposed to function as an interface in the context of algorithmically controlled guidance processes.\u2019<\/span> Based as they are on numeric values, insular procedures, and a series of recursive instructions, we can therefore understand \u2018operational images\u2019 in algorithmic terms. Furthermore, the innate purposiveness of \u2018operational images\u2019 in contemporary theatres of warfare and guidance systems is, despite the opacities involved in their processes, repeatedly revealed in their real world impact. The extent to which they are deployed in surveillance technologies and the establishment of autonomous models of warfare \u2013 defined here as \u2018algorithmically controlled guidance processes\u2019 \u2013 ensures that \u2018operational images\u2019 are used to target and, indeed, kill people.<\/span><\/p>\n Through locating the epistemological and actual violence that impacts communities and individuals who are captured, or \u2018tagged\u2019, by autonomous systems, we can further reveal the extent to which the legacy of colonialism informs the algorithmic logic of neo-colonial imperialism. The logistics of data extraction, not to mention the violence perpetuated as a result of such activities, is all too amply captured in Aim\u00e9 C\u00e9saire\u2019s succinct phrase: \u2018colonisation = thingification\u2019. Through this resonant formulation, C\u00e9saire highlights both the inherent processes of dehumanization practised by colonial powers and how, in turn, this produced the docile and productive \u2013 that is, passive and commodified \u2013 body of the colonized.<\/span><\/p>\n As befits his time, C\u00e9saire understood these configurations primarily in terms of wealth extraction (raw materials) and the exploitation of physical, indentured labour. However, his thesis is also prescient in its understanding of how colonization seeks unmitigated control over the future, if only to pre-empt and extinguish elements that did not accord with the avowed aims and priorities of imperialism: \u2018I am talking about societies drained of their essence, cultures trampled underfoot, institutions undermined, lands confiscated, religions smashed, magnificent artistic creations destroyed, extraordinary <\/span>possibilities<\/span><\/i> wiped out.\u2019<\/span> The exploitation of raw materials, labour and people, realized through the violent projections of western knowledge and power, employed a process of dehumanization that deferred, if not truncated, the quantum possibilities of <\/span>future<\/span><\/i> realities.\u00a0<\/span><\/p>\n In the context of the Middle East, the management of risk and threat prediction \u2013\u00a0the <\/span>containment<\/span><\/i> of the future \u2013 is profoundly reliant on the deployment of machine learning and computer vision, a fact that was already apparent in 2003 when, in the lead up to the invasion of Iraq, George W. Bush announced that \u2018if we wait for threats to fully materialize, we will have waited too long.\u2019<\/span> Implied in Bush\u2019s statement, whether he intended it or not, was the unspoken assumption that counter-terrorism would be necessarily aided by semi if not fully-autonomous weapons systems capable of maintaining and supporting the military strategy of anticipatory and preventative self-defence.<\/span><\/p>\n To predict threat, this logic goes, you have to see further than the human eye and act quicker than the human brain; to pre-empt threat you have to be ready to determine and exclude (eradicate) the \u2018unknown unknowns\u2019. <\/span>Although it has been a historical mainstay of military tactics, the use of pre-emptive, or anticipatory, self-defence \u2013 the so-called \u2018Bush doctrine\u2019 \u2013 is today seen as a dubious legacy of the attacks on the US on 11 September 2001. Despite the absence of any evidence related to Iraqi involvement in the events of 9\/11, the invasion of Iraq in 2003 \u2013 to take but one particularly egregious example \u2013 was a pre-emptive war waged by the US and its erstwhile allies in order to mitigate against such attacks in the future.<\/span><\/p>\n Harun Farocki, Eye Machine III, \u00a9 Harun Farocki, 2003. Image via Springerin<\/p>\n<\/div>\n In keeping with the ambition to predict \u2018unknown unknowns\u2019, Alex Karp, the CEO of Palantir, wrote an opinion piece for <\/span>The New York Times <\/span><\/i>in July 2023. Published 20 years after the invasion of Iraq, and therefore written in a different era, the apparent threats to US security and the need for robust methods of pre-emptive warfare, were in the forefront of Karp\u2019s thinking, nowhere more so than when he espoused the seemingly prophetic if not oracle-like capacities of AI predictive systems.<\/span><\/p>\n Conceding that the use of AI in contemporary warfare needs to be carefully monitored and regulated, he proposed that those involved in overseeing such checks and balances \u2013 including Palantir, the US government, the US military and other industry-wide bodies \u2013 face a choice similar to the one the world confronted in the 1940s. \u2018The choice we face is whether to rein in or even halt the development of the most advanced forms of artificial intelligence, which some argue may threaten or someday supersede humanity, or to allow more unfettered experimentation with a technology that has the potential to shape the international politics of this century in the way nuclear arms shaped the last one.\u2019<\/span><\/p>\n Admitting that the most recent versions of AI, including the so-called Large Language Models (LLMs) that have become increasingly popular in machine learning, are impossible to understand for user and programmer alike, Karp accepted that what \u2018has emerged from that trillion-dimensional space is opaque and mysterious\u2019.<\/span> It would nevertheless appear that the \u2018known unknowns\u2019 of AI, the professed opacity of its operative logic (not to mention the demonstrable inclination towards erroneous prediction, or hallucinations), can nevertheless predict the \u2018unknown unknowns\u2019 associated with the forecasting of threat, at least in the sphere of the predictive analytics championed by Palantir. <\/span>Perceiving this quandary and asserting, without much by way of detail, that it will be essential to \u2018allow more seamless collaboration between human operators and their algorithmic counterparts, to ensure that the machine remains subordinate to its creator\u2019, Karp\u2019s overall argument is that we must not \u2018shy away from building sharp tools for fear they may be turned against us\u2019.<\/span>\u00a0<\/span><\/p>\n This summary of the continued dilemmas in the applications of AI systems in warfare, including the peril of machines that turn on us, needs to be taken seriously insofar as Karp is one of the few people who can talk, in his capacity as the CEO of Palantir, with an insider\u2019s insight into their future deployment. Widely seen as the leading proponent of predictive analytics in warfare, Palantir seldom hesitates when it comes to advocating the expansion of AI technologies in contemporary theatres of war, policing, information management and data analytics more broadly.<\/span><\/p>\n In tune with its avowed ambition to see AI more fully incorporated into theatres of war, its website is forthright on this matter. We learn, for example, that \u2018new aviation modernization efforts extend the reach of Army intelligence, manpower and equipment to dynamically deter the threat at <\/span>extended range<\/span><\/i>. At Palantir, we deploy AI\/ML-enabled solutions onto airborne platforms so that users can<\/span> see farther<\/span><\/i>, generate insights faster and react at the speed of relevance.\u2019 <\/span>As to what reacting \u2018at the speed of relevance\u2019 means we can only surmise this has to do with the pre-emptive martial logic of autonomously anticipating and eradicating threat <\/span>before<\/span><\/i> it becomes manifest.<\/span><\/p>\n Palantir\u2019s stated objective to produce predictive models and AI solutions that enable military planners to (autonomously or otherwise) \u2018see farther\u2019 is not only ample corroboration of its reliance on the inferential, or predictive, qualities of AI but, given its ascendant position in relation to the US government and the Pentagon, a clear indication of how such neo-colonial technologies will determine the prosecution and outcomes of future wars in the Middle East.<\/span><\/p>\n This ambition to \u2018see farther\u2019, already manifest in colonial technologies of mapping, also supports the neo-colonial ambition to see that which <\/span>cannot be seen<\/span><\/i> \u2013 or that which can only be seen <\/span>through <\/span><\/i>the algorithmic gaze and its rationalization of future realities. As Edward Said argues in his seminal volume <\/span>Orientalism<\/span><\/i>, the function of the imperial gaze \u2013\u00a0and colonial discourse more broadly \u2013 was \u2018to divide, deploy, schematize, tabulate, index, and record everything in sight<\/span> (and <\/span>out of sight<\/span><\/i>)\u2019.<\/span>\u00a0This is the future-oriented algorithmic \u2018vision\u2019 of a neo-colonial world order \u2013 an order maintained and supported by AI apparatuses that seek to quarter, appropriate, realign, predict and record everything in sight \u2013 and, critically, everything <\/span>out of sight.<\/span><\/i>\u00a0<\/span><\/p>\n Although routinely presented as an objective \u2018view from nowhere\u2019 (a strategy used in colonial cartography), AI-powered models of unmanned aerial surveillance and autonomous weapons systems \u2013 given the enthusiastic emphasis on extrapolation and prediction \u2013 are epistemic structures that <\/span>produce<\/span><\/i> realities. These computational structures, provoking as they do actual events in the world, can also be used to justify the event of real violence.<\/span> For all the apparent viability, not to mention questionable validity, of the AI-powered image processing models deployed across the Middle East, we need to therefore observe the degree to which \u2018algorithms are political in the sense that they help to make the world appear in certain ways rather than others.<\/span><\/p>\n Carmel of southern Palestine, photographed between 1950 and 1977, Matson (G. Eric and Edith) Photograph Collection, Library of Congress. Image via Springerin<\/p>\n<\/div>\n Speaking of algorithmic politics in this sense, then, refers to the idea that realities are never given but <\/span>brought into being<\/span><\/i> and <\/span>actualized<\/span><\/i> in and through algorithmic systems.\u2019<\/span> This is to recall that colonization, as per Said\u2019s persuasive insights, was a \u2018systematic discipline by which European culture was able to manage \u2013\u00a0and even <\/span>produce<\/span><\/i> \u2013 the Orient politically, socially, militarily, ideologically, scientifically, and imaginatively during the post-Enlightenment period.\u2019<\/span> The fact that Said\u2019s insights have become largely accepted if not conventional should not distract us from the fact that the age of AI has witnessed an insidious re-inscription of the racial, ethnic and social determinism that figured throughout imperial ventures and their enthusiastic support for colonialism.\u00a0<\/span><\/p>\n In the milieu of so-called Big Data, machine learning, data scraping and applied algorithms, a form of digital imperialism is being profoundly, not to mention lucratively, programmed into neo-colonial prototypes of drone reconnaissance, satellite surveillance and autonomous forms of warfare, nowhere more so than in the Middle East, a nebulous, often politically concocted, region that has long been a testing ground for Western technologies.<\/span> In suggesting that the machinic \u2018eye\u2019, the \u2018I\u2019 associated with cartographic and other methods of mapping, has evolved into an unaccountable, <\/span>detached<\/span><\/i> algorithmic gaze is to highlight, finally, a further distinction: the devolution of deliberative, ocular-centric models of seeing and thinking to the recursive realm of algorithms reveals the callous rendering of subjects in terms of their disposability or replaceability, the latter being a key feature \u2013\u00a0as observed by C\u00e9saire \u2013 of colonial discourse and practice.<\/span><\/p>\n In light of these computational complicities and algorithmic anxieties, the <\/span>detached<\/span><\/i> apparatuses of neo-colonization, we might want to ask whether there is a correlation between automation and the disavowal of culpability: does the deferral of perception, and the decision-making processes we associate with the ocular-centric domain of human vision, to autonomous apparatuses guarantee that we correspondingly reject legal, political and individual responsibility for the applications of AI? Has AI, in sum, become an alibi \u2013 a means to disavow individual, martial and governmental liability when it comes to algorithmic determinations of life and, indeed, death?<\/span><\/p>\n<\/p><\/div>\n![\"\"](\"https:\/\/www.eurozine.com\/wp-content\/uploads\/2024\/08\/Eye-Machine_III_A.0009-1024x819.jpg\")
<\/p>\nProjecting \u2018thingification\u2019<\/h2>\n
![\"\"](\"https:\/\/www.eurozine.com\/wp-content\/uploads\/2024\/08\/Eye-Machine_III_A.0010-1024x819.jpg\")
<\/p>\nPredicting \u2018unknown unknowns\u2019<\/b><\/h2>\n
![\"\"](\"https:\/\/www.eurozine.com\/wp-content\/uploads\/2024\/08\/Eye_Machine-III-dch_A.0005-1024x819.jpg\")
<\/p>\nThe AI alibi<\/h2>\n
![\"\"](\"https:\/\/www.eurozine.com\/wp-content\/uploads\/2024\/08\/4.jpg\")
<\/p>\n