In 1932 Andrei Kolmogorov, under intense pressure from the KGB for his alleged gay relationships with Pavel Aleksandrov, formalized modern probability theory into an engine of radical contingency, thereby ushering in the current era of probabilistic artificial intelligence marked by „hallucinating“ large language models and Bayesian networks. In close reading of the tumultous Soviet history of media and mathematics, this paper offer a contribution to the ongoing theorization of environment media with a few profiles drawn from my current book on the Soviet prehistory of artificial intelligence. In particular, it offers a review of what I am calling „sreda theory,“ or a Russian-language media theory of environments. Snatches of the concept of sreda, or the Russian word for „environment,“ „Wednesday,“ and a near synonym with „medium,“ emerges across a tumultuous early twentieth century of Soviet intellectuals. In addition to the probabilistic fields of Kolmogorov, the notion of sreda emerges centrally in the grounded, yet cosmically writings of Theodosius Dobzhansky on the modern synthesis, Vladimir Vernadsky on the biosphere, and then Yekateryna Yushchenko’s conceptualization of „addressatsia“ (a decade before „pointers“) in early 1950s Kyiv. Ending in radio-controlled robots navigating the aftermath of Chernobyl, this talk traces out a (particularly Soviet-Ukrainian) history of media environments as smart environments, in which artificial intelligence materializes into an uncanny environment. What makes the noosphere, the sets of modern probability theory, the dynamism of statistical population models, and the lethally irradiated reactor four uncanny except their mutual, ambiguous question of life: what relationship, if any, do smart environments have with life as we know it? How might environmental media theorists, with the tragic Soviet annals of smart environments in hand survey our own era of data-driven surveillance, better critically reclaim and re-envision our own current era of coal-powered artificial intelligence amid the breakneck escalation of climate crises? The paper will conclude with remarks on these and other questions.

About the lecturer

Benjamin Peters is the Mercator Fellow at the University of Siegen in the Fall 2024 as well as the Hazel Rogers Associate Professor of Media Studies at the University of Tulsa (a sister university of the University of Siegen), where he also holds faculty affiliation with the Cyber School and Honors Program. He is also as well as a faculty affiliate at the Information Society Project at Yale Law School. He is currently at work on three book projects: a Soviet prehistory to artificial intelligence from which this talk draws, a cultural history of Russian hackers coauthored with Marijeta Bozovic (Yale Slavic), and a short book of letters to his college-aged child called How to College.

All events take place in hybrid form (on site and via Webex). If you would like to attend on site, no registration is required. To attend the lecture online via Webex, please register here.

About the lecture series