Response to The /nstitute for Other /ntelligences

Mashinka Firunts Hakopian South Pasadena: X Artists Books, 2022

/nstitute for Other /ntelligences, (image by permission of the author).

Mashinka Firunts Hakopian presents The Institute for Other Intelligences as a report from the future. Early on the text makes clear that one of its purposes is to offer a warning similar to that posted by the US Department of Energy on the Waste Isolation Pilot Plant in New Mexico. The sign on that disposal site is intended to inform generations far into the future about the radioactive waste and toxins buried there and their capacity to contaminate. Keeping future contamination under control depends on actions in the present. The toxins targeted for containment in Hakopian’s text are not chemicals. Instead, they are breaches of data ethics whose capacity to produced damage in the socio-techno ecosystem aligns with practices of racism, gender bias, and other forms of discrimination linked to demographic markers and related concerns. The long and longer term consequences of actions that may arise from  systems with these built-in and/or self-reinforcing prejudices are unpredictable, but their impact in the current environment is sufficiently pernicious to warrant deliberate intervention. These are familiar—but largely ignored or unheeded–warnings as data practices continue to exceed regulations and cautions of all kinds while reinforcing lines of exclusion and abuse in the social sphere.

            Messages from the future are a familiar trope of science-fiction. In The Terminator (1984), the time-travelling assassin, played by Arnold Schwarzenegger, is sent to eliminate a future hero-savior who would lead the humans to triumph in the coming war against evil machines. This is a classic move, an intervention in the sequence of historical events triggered by altering the past with knowledge from the future (e.g. going back in time to eliminate Adolf Hitler). Other standards in the genre involve passing through wormholes in warped space-time travel or being captured and kept as curiosities by advanced cultures who see humans as undeveloped but interesting specimens (i.e. pets). Paradoxically, all messages from the future are actually from the past by the time they arrive, like light from distant stars. Of course, other  equally disturbing science fiction accounts include the dystopic visions imagining the future ahead, among which George Orwell’s 1984 remains altogether too terrifyingly prescient. Whether imagined in either direction—from or to the future—dystopic scenarios that draw on actual social conditions serve as cautionary tales. Some propose alternatives, like this one by Hakopian, which advocates for intervention in specific ways.

            The volume is modelled like a conference proceedings, a conceit that works well for organizing a combination of statements, exercises, a keynote talk and other materials all presented in this stylishly designed format. The reports and proceedings from the Other Intelligences are translated from non-linguistic machine communication into English for this readership (or so the “translator” informs us). Time stamps have been reformatted from the nano-second scale that indicates simultaneity into a sequential ones for human readers. These and other touches provide “truthy” frames throughout the book. For instance, one section notes that many pernicious technologies have, by the time of the future writing of this work, been obsolete for some time. These include “facial recognition, risk assessment scoring, automated hiring, predictive policing, and others.” (13) The major concern of the Institute is/was/will be to “unsettle” the assumptions on which AI training was proceeding. Chiefly, this means recognizing the fallacy that data-driven approaches to intelligence would lead to objective decision-making in supposedly value-neutral judgment terms.

These discussions have long been familiar staples in the discourse of critical data studies—as well as those science fiction worlds already noted where humans are locked into apocalyptic struggles for survival with robots, systems, and automated entities or instruments gone radically wrong. By contrast, the Intelligences of the Institute are focused on correcting the social injustice of the present that is enabled by algorithmic bias and mis-directed use of so-called intelligent systems. To achieve this, the Institute has the mission of training oppositional automata, those that take their impetus from response to current problems linked to AI systems and their instrumentalization.  

            Current trends in critical data studies form the intellectual framework in this book and the Other Intelligences have been very busy keeping up with the literature in the field, doing their homework with the diligence of eager seminar students. The reading list is packed with familiar names and references: Wendy Hui Kyong Chun, Catherine D’Ignazio and Lauren Klein, Stephanie Dinkins, and Safiya Noble, are among others who work at the intersection of social justice, algorithmic bias, and activism. The recommendations for transformative activities are equally familiar in their attempts at undercutting traditional (read western, masculinist, hegemonic) assumptions about knowledge such as “privileged claims” about “certainty,” “forms of objectification,” and “presumed impartiality.” These critical arguments have a lineage in feminist theory as well as the larger history of dismantling Cartesian thought and Rationalism more generally.

But what is at stake here is the commitment to encoding a future from an alternate perspective. In the present, certain populations and demographic groups are targeted explicitly by automated systems. Identifying how these practices work to reinforce circumstances of inequity and social injustice is a necessary intervention. Resetting future developments requires rethinking our own processes as well as those we have enabled in automata. In the “Keynote” portion of the presentation, work by Maya Ganesh, A is for Another, serves to provoke a way of thinking that points “outward beyond its own systems of meaning.” As the text moves towards its conclusion, the references keep bringing in discussions of chat bots and missing data sets, proposals for intelligent systems that appear refuse the commands of humans, and counter-cultural value systems. But is this sufficient–even if it is the necessary start point? A degree of predictability emerges from the pattern of references, the “good” alternatives are consistently put in opposition to the premises identified as oppressive and unjust. All well and good–massive correction is certainly needed in artificial systems working perniciously to reinforce abusive aspects of existing power structures.

But what does thinking beyond familiar habits actually mean? The ethical commitment of these future intelligences is impressive, but could they be pushed towards a more imaginative invention? I was a bit surprised, for instance, to find that none of the materials cited in the bibliographies had a publication date much later than 2021. In all seriousness, I would like to imagine their vocabularies, approaches, methods, and topics will shift in the unspecified “future” to which they belong. What would be on the reading list in 3021? Why not conjure reports from future millennia on “reparative data filtration,” “CRISPR techniques for bias correction in organic computing systems,” “quantum tunneling for algorithmic anticipation modeling,” “outlook engineering,” and “flexible justice systems”? In other words, what might be contained in reports and white papers from the research groups of the Other Intelligences themselves that goes beyond current orthodoxies and formulations? I want a glimpse into future concepts, critical paradigms beyond those of the present, examples of the liberatory transformation of the very terms on which intelligence and its applications, instrumentalizations, and possibilities are imagined. In effect, how do we think outside of reactive opposition? Perhaps the next volume of this work could be focused on such a bibliographical reference list so that we might shift beyond the usual critiques, important as they are, to an as-yet unarticulated model developed from fully alternative premises. What will the “future” think of us? Much depends on who is doing the thinking, what entity and what concept of intelligence—or thought—is in play. Imagination is a crucial tool of survival. More ahead, thanks to this provocation.