The Global AI Narratives team, along with local partner Teplitsa of Social Technologies, convened a workshop in Saint Petersburg, Russia, on the 12th and 13th of September 2019. The event was designed to explore Russian narratives and perceptions of artificial intelligence, and consisted of an evening public lecture followed by a full-day workshop event. The two day event attracted AI stakeholders from across Russia and some from further afield, as well as interested members of the public, many of whom were local to Saint Petersburg.
Workshop Participants - photograph courtesy of Anna Balakhontseva
Public Lecture
The sold-out public lecture was designed to act as a setting for people with an interest in AI to engage in the narratives that surround this popular and fascinating topic. In addition to providing a forum to begin creating a dialogue around Russian AI narratives and their global counterparts, this event was also designed to offer AI stakeholders and interested members of the public a platform to engage with each other. With the intention of building dialogue between cultures, we were pleased to foster a multi-lingual public lecture event; one keynote lecture was presented in English, and the other in Russian, with the panel discussion accommodating both languages (ably facilitated by some excellent simultaneous translation). Dr Kanta Dihal opened the public event with a lecture entitled ‘Golden Maidens and Killer Robots: Which Stories have Shaped Western AI?’ which explored the history of the way that humans think about and narrate ideas of intelligence.
The second keynote speaker, Anton Pervushin is an acclaimed science fiction author, editor, judge and journalist, and he offered an intriguing look at the history of Russian science fiction and the role of humanity and the soul in these narratives. Historically Russian society (and to some extent science fiction) has focussed on mechanical or industrial-style robots, such as the recently viral Russian robot Skybot F850 (find the official Twitter here). In fiction, narratives around this type of robot often focus on the inability of the robot to pass as human; many focus on the importance of teaching AIs unambiguously human traits, such as feeling emotions and drinking vodka. In this respect the robot character Elektronic represents a traditional tale; the robot boy becomes self-aware and decides to become part of human society. In the beginning he is a poor imitation of a real human, but quickly learns and eventually starts to become emotionally attached to his ‘human’ life. Anton argued that science fiction narratives suggest that without the step towards emotional engagement with the world around them, AI can only ever be a poor imitation of humanity; to become truly intelligent, the AI must become human.
These lectures were followed by a panel discussion with Dr Dihal, Anton Pervushin, Alexey Sidorenko (Teplitsa), and Tonii Leach (GAIN). The panel discussed the influence that golden-age science fiction was still having on both current science fiction writers and the public more widely, the difference between western and other global AI narratives, and good examples of AI in contemporary science fiction. The panel was also pleased to discuss a range of interesting and topical questions from the audience, including the potential for a positive or negative AI future, and whether Google, Siri, or Alexa is the best AI personal assistant - not only could the panel not agree, but neither could the audience!
Public Lecture Discussion Panel: (L-R) Alexey Sidorenko, Tonii Leach, Anton Pervushin, Botty McBottface, Kanta Dihal
Workshop
The first session entitled ‘The History of AI’ featured presentations from Anna Vlasova, Anton Pervushin and Stephen Cave. Anna Vlasova discussed the changing expectations of the Russian public regarding customer service chatbots, describing three distinct periods of public expectation starting with the mid-2000s, which was heavily influenced by dystopian science fiction narratives, and growing more realistic since then with improved public understanding of the capability of AI technology. Anton Pervushin explored the history of the robot as the main concept of AI in Soviet science fiction, from the political revolutionary narratives such as Karel Čapek’s ‘R.U.R’ in the pre-Second World War era, to the dystopian narratives of hyper-intelligent military robot technology in the post-Second World War context. Following this, and the success of Isaac Asimov’s science fiction in the mid-1960s, there was a steep decline in native AI science fiction, which has been compounded by the influx of Western AI narratives from Hollywood films and TV shows. Stephen Cave spoke about how we conceptualise the idea of intelligence, and the historical bases for these ideas.
The second session was entitled ‘AI and Society’, and featured presentations from Alexey Sidorenko, Natalia Tregubova, Marina Shilina and Tonii Leach. Alexey Sidorenko addressed the narratives around unemployment that are prevalent in the Russian media, describing three key narratives: AI as a ‘thief’ which will steal jobs, AI as a ‘servant’ or tool to undertake jobs humans do not wish to do, and AI as an ‘instigator’ which will act as a catalyst for humans to rebel against other humans. Natalia Tregubova described issues of AI mediating human interaction in the digital age, and asked whether AI shifting from a goal-orientated technology towards an interaction-orientated technology might further disrupt traditional human-human interaction channels. Marina Shilina discussed how and why the Russian people might develop trust in AI technology, and whether the Russian government’s digitisation model is driving adoption of AI technology in industry and promoting trust in these technologies at a societal level. Tonii Leach addressed the role of science fiction narratives in global AI policy, and how these narratives frame the debate around discrimination in AI.
Workshop Session Q&A - photograph courtesy of Anna Balakhontseva
The third session was entitled ‘Public Perceptions of AI’, and featured presentations from Ilya Ovchinnikov, Liliia Zemnukhova, Kanta Dihal and Irina Kotkina. Ilya Ovchinnikov described the variety of narratives of customer engagement with AI support services, and some of the challenges for industry associated with gauging human reactions to AI technology. Liliia Zemnukhova presented the results of a survey on Russian public opinion on AI, and discussed how these results might impact on the development of AI in Russia, particularly with a focus on building upon an overall positive perception of AI held by the general public. Kanta Dihal was able to offer a counterpoint to Liliia’s presentation by discussing the results of a survey of British perceptions of AI, and how the overarching perceptions of AI in Britain tend to be negative. Irina Kotkina discussed how narratives around fear of AI are perpetuated, suggesting these fears have developed in Russia from the fear of extraterrestrials in the mid-late twentieth century, which has been compounded by the rise of conspiracy theory culture online.
The fourth session was entitled ‘AI and Politics in Russia’, and featured presentations from Andrey Rezaev, Anzhelika Solovyeva, Alena Popova and Elena Shakhova. Andrey Rezaev talked about the requirement to move away from interdisciplinarity in AI research going forward. Anzhelika Solovyeva discussed the strategic culture of Russian military development and how this has impacted upon the development and adoption of military AI technology. Alena Popova addressed the role of facial recognition technology in programmes of state monitoring and suppression, and how data driven surveillance can act to limit the rights of individuals if the concepts of necessity, proportionality, and transparency are not adequately embedded in the development of the system. Elena Shakhova described some of the possible pitfalls of introducing AI technology into judicial systems, as well as the potential for AI to be used to improve the objectivity of decisions made within the court system.
The workshop concluded with a breakout session, where all participants engaged in group activities to try to answer some pressing questions around AI in Russia based on the variety of interesting presentations given during the workshop. Ideas discussed included similarities between Russian fears around AI and fears in the West, including automation and job replacement, privacy and the role of AI in politics. Against this backdrop, however, the participants felt strongly that the specific response to AI in Russia is also deeply embedded in the political and social history of the country, and that this provides for the Russian public to engage with AI in a way that Western populations do not; in particular, the specifically Russian history of AI narratives such as Elektronic, provided little cross-over with Western narratives in the formative years of science fiction focussing on AI and thus created a very different reception to AI amongst the Russian population. However, it was also considered that the recent availability of Western AI narratives has started to change not only public perception of AI in Russia today, but is also beginning to influence the kind of narratives produced within Russia, from fiction, to science writing and journalism - it seems that Western narratives have truly arrived in Russia.
Worship Discussion - Photograph courtesy of Anna Balakhontseva
The CFI team are working on a more detailed report of this meeting, and we look forward to collaborating further with our wonderful Russian partners. In the meantime, this is a summary of the clearest points to emerge. Russian narratives around AI are:
Generally much more positive about AI
Much more utilitarian in considering AI as a tool
influenced by a different historical tradition, which emphasises Russian and Soviet identities
concerned with maintaining the Russian position in global AI standings
heavily focussed on the implications of state control of AI technology
A full report on the GAIN Russia 2019 workshop will be produced shortly.