Welcome to the Digital Humanities / Artificial Intelligence Seminar!
Fostered by the creation of new algorithms, computation power and the development of deep learning techniques, Artificial Intelligence needs constantly to confront new issues and data sets in order to deepen its methodologies and increase its range of scientific applications. Digital humanities, developing digital science methodologies in the study of humanities and using the critical approaches of humanities in the analysis of the contemporary “digital revolutions”, are constantly in search of new tools to explore more and more complex and diversified data sets.
The coupling AI/DH is globally emerging as one key interface for both domains and will probably prove to be a deep transformative trend in tomorrow intellectual world.
The ambition of this seminar is to be one of the places where this coupling is shaped, fostered and analyzed. It intends to offer a forum where both communities, understood in a very inclusive way, exchange around emerging issues, ongoing projects, and past experiences in order to build a common language, a shared space, and to encourage innovative cooperation on the long run.
You can access here the list of past seminars.
June 4, 2020, 15h-17h, room online (link here, ID meeting : 981 5273 8608).
Sietske Fransen & Leonardo Impett (Max-Planck-Institut für Kunstgeschichte)
Title: Print, Code, Data: New Media Disruptions and Scientific Visualization
Abstract: This paper discusses changes in scientific diagramming in response to new media disruptions: the printing press, and online data/research code. In the first case, the role of handwritten documents and the visual forms of scientific diagramming re-align in response to the circulational economics and medial accessibility of the printing press in early modern Europe. In the second, published research code unsettles the principle, common in the second half of the twentieth century, that a peer-reviewed article in computer science ought to outline its methods with enough detail to enable repeatability.
The printing press brought benefits as well as restrictions to the inclusion of diagrams in scientific works. Some of the downsides were that not every printer was able to manufacture separate wood blocks and/or copper plates that could contain the diagram as if hand-drawn. Instead, diagrams were often made entirely out of typeface. On the other hand, the quick spread of the use of printed books in addition to manuscripts, opened new roles for the manuscript as a medium of creativity. In the early days of print, it is therefore in manuscripts that we can find the visualization of scientific processes, which form the background to printed material.
The information sufficient for 'replicability' in computer science (which in the physical sciences has meant 'formal experimental methodology', but in computer science is epistemically closer to the research itself) had most often been included in tables, schematizations and heavily-labelled diagrams, sometimes augmented by so-called 'pseudocode' (a kind of software caricature, which cannot itself be run on a machine). The inclusion of research code thus dramatically displaces the role of scientific diagrams in machine learning research: from a notational system which ideally contains sufficient information to reproduce an algorithm (akin to electrical circuit diagrams) to a didactic visualization technique (as in schoolbook diagrams of the Carbon Cycle). In Badiou's (1968) terminology, diagrams shift from symbolic formal systems to synthetic spatializations of non-spatial processes. The relationship between 'research output' (as the commodity produced by computer-science research groups) and its constituent components (text, diagram, code, data) is further destabilized by deep learning techniques (which rely on vast amounts of training data) : no longer are algorithms published on their own, but rather trained models, assemblages of both data and software, again shifting the onus of reproducibility (and, therefore, the function of scientific notation). The changed epistemological role of neural network visualizations allows for a far greater formal instability, leading to the rich ecology of visual solutions (Alexnet, VGG, DeepFace) to the problem of notating multidimensional neural network architectures.
By comparing the impact of new media on the use, form and distribution of diagrams in the early modern period, with the impact of code on the role of diagrams in computer science publications, we are opening up a conversation about the influence of new media on science, both in history and in current practice.
June 8, 2020, 12h-14h, room online (link here).
Antonio Casilli (Paris School of Telecommunications (Telecom Paris))
Title: The last mile of inequality: What COVID-19 is doing to labor and automation
Abstract: The ongoing COVID-19 crisis, with it lockdowns, mass unemployment, and increased health risks, has been described as a automation-forcing event, poised to accelerate the introduction of automated processes replacing human workers. Nevertheless, a growing body of literature has emphasized the human contribution to machine learning. Especially platform-based digital labor performed by global crowds of underpaid micro-workers or extracting data from cab-hailing drivers and bike couriers, turns out to play an crucial role. Although the pandemic has been regarded as the triumph of 'smart work', telecommuting during periods of lockdown and closures concern only about 25 percents of workers. A class gradient seems to be at play, as platform-assisted telework is common among higher-income brackets, while people on lower rungs of the income ladder are more likely to hold jobs that involve physical proximity, which are deemed essential and cannot be moved online or interrupted. These include two groups of contingent workers performing what can be described as 'the last mile of logistics' (delivery, driving, maintenance and other gigs at the end of the supply chain) and the 'last mile of automation' (human-in-the-loop tasks such as data preparation, content moderation and algorithm verification). Indeed during lockdown, both logistic and micro-work platforms have reported a rise in activity – with millions signing up to be couriers, drivers, moderators, data trainers. The COVID-19 pandemic has thus given unprecedented visibility to these workers, but without increased social security. Their activities are equally carried out in public spaces, in offices, or from home—yet they generally expose workers to higher health risks with poor pay, no insurance, and no sick leave. Last mile platform workers shoulder a disproportionate share of the risk associated with ensuring economic continuity. Emerging scenarios include use of industrial actions to increase recognition and improve their working conditions. COVID-19 has opened spaces of visibility by organizing workers across Europe, South America, and the US. Since March 2020, Instacart walkouts, Glovo and Deliveroo street rallies, Amazon 'virtual walkouts' have started demanding health measures or protesting remuneration cuts.