I managed to catch a screening of the new Shalini Kantayya documentary, Coded Bias, through EDRi. It tells the story of Joy Bualomwini‘s discovery of systematic discrepancies in the performance of algorithms across races and genders. The tone was lively and accessible, with a good tempo, and the cast of characters presented did a good job showcasing a cross-section of female voices in the tech policy space. It was particularly good to see several authors that appear on my syllabus, such as Cathy O’Neil, Zeynep Tufekci, and Virginia Eubanks.
Interesting article in the MIT Tech Review (via /.) detailing research performed at Northwestern University (paper on ArXiv) on how potentially to leverage the power of collective action in order to counter pervasive data collection strategies by internet companies. Three such methods are discussed: data strikes (refusal to use data-invasive services), data poisoning (providing false and misleading data), and conscious data contribution (to privacy-respecting competitors).
Conscious data contribution and data strikes are relatively straightforward Aventine secessions, but depend decisively on the availability of alternative services (or the acceptability of degraded performance for the mobilized users on less-than-perfect substitutes).
The effectiveness of data poisoning, on the other hand, turns on the type of surveillance one is trying to stifle (as I have argued in I labirinti). If material efficacy is at stake, it can be decisive (e.g., faulty info can make a terrorist manhunt fail). Unsurprisingly, this type of strategic disinformation has featured in the plot of many works of fiction, both featuring and not featuring AIs. But if what’s at stake is the perception of efficacy, data poisoning is only an effective counterstrategy inasmuch as it destroys the legitimacy of the decisions made on the basis of the collected data (at what point, for instance, do advertisers stop working with Google because its database is irrevocably compromised?). In some cases of AI/ML adoption, in which the offloading of responsibility and the containment of costs are the foremost goals, there already is very broad tolerance for bias (i.e., faulty training data).
Hence in general the fix is not exclusively technical: political mobilization must be activated to cash in on the contradictions these data activism interventions bring to light.
An interesting report in Medium (via /.) discusses the PRC’s new pervasive surveillance program, Sharp Eyes. The program, which complements several other mass surveillance initiatives by the Chinese government, such as SkyNet, is aimed especially at rural communities and small towns. With all the caveats related to the fragmentary nature of the information available to outside researchers, it appears that Sharp Eyes’ main characteristic is being community-driven: the feeds from CCTV cameras monitoring public spaces are made accessible to individuals in the community, whether at home from their TVs and monitors or through smartphone apps. Hence, local communities become responsible for monitoring themselves (and providing denunciations of deviants to the authorities).
This outsourcing of social control is clearly a labor-saving initiative, which itself ties in to a long-run, classic theme in Chinese governance. It is not hard to perceive how such a scheme may encourage social homogeneization and irregimentation dynamics, and be especially effective against stigmatized minorities. After all, the entire system of Chinese official surveillance is more or less formally linked to the controversial Social Credit System, a scoring of the population for ideological and financial conformity.
However, I wonder whether a community-driven surveillance program, in rendering society more transparent to itself, does not also potentially offer accountability tools to civil society vis-à-vis the government. After all, complete visibility of public space by all members of society also can mean exposure and documentation of specific public instances of abuse of authority, such as police brutality. Such cases could of course be blacked out of the feeds, but such a heavy-handed tactic would cut into the propaganda value of the transparency initiative and affect public trust in the system. Alternatively, offending material could be removed more seamlessly through deep fake interventions, but the resources necessary for such a level of tampering, including the additional layer of bureaucracy needed to curate live feeds, would seem ultimately self-defeating in terms of the cost-cutting rationale.
In any case, including the monitored public within the monitoring loop (and emphasizing the collective responsibility aspect of the practice over the atomizing, pervasive-suspicion one) promises to create novel practical and theoretical challenges for mass surveillance.
Vice reports on a Tokyo-based company, DeepScore, pitching software for the automatic recognition of ‘trustworthiness’, e.g. in loan applicants. Although their claimed false-negative rate of 30% may not sound particularly impressive, it must of course be compared to well-known human biases in lending decisions. Perhaps more interesting is the instrumentalization cycle, which is all but assured to take place if DeepScore’s algorithm gains wide acceptance. On the one hand, the algorithm’s goal is to create a precise definition for a broad and vague human characteristic like trustworthiness—that is to say, to operationalize it. Then, if the algorithm is successful on its training sample and becomes adopted by real-world decision-makers, the social power of the adopters reifies the research hypothesis: trustworthiness becomes what the algorithm says it is (because money talks). Thus, the behavioral redefinition of a folk psychology concept comes to fruition. On the other hand, however, instrumentalization immediately kicks in, as users attempt to game the operationalized definition, by managing to present the algorithmically-approved symptoms without the underlying condition (sincerity). Hence, the signal loses strength, and the cycle completes. The fact that DeepScore’s trustworthiness algorithm is intended for credit markets in South-East Asia, where there exist populations without access to traditional credit-scoring channels, merely clarifies the ‘predatory inclusion’ logic of such practices (v. supra).
Several reports are circulating (e.g., via /.) of a court case in New Jersey in which the defendant won the right to audit proprietary genetic testing software for errors or potential sources of bias. It being a murder trial, this is about as close to a life-or-death use-case as possible.
Given the stakes, it is understandable that a low-trust standard should prevail in forensic matters, rendering an audit indispensable (nor is the firm’s “complexity defence” anything short of untenable). What is surprising, rather, is how long it took to obtain this type of judicial precedent. The authoritativeness deficit of algorithms is a topic of burning intensity generally; that in such a failure-critical area a business model based on proprietary secrecy has managed to survive is truly remarkable. It is safe to say that this challenge will hardly be the last. Ultimately, freely auditable software would seem to be the superior systemic answer for this type of applications.