Working group conveners: Docent Salla-Maaria Laaksonen, University of Helsinki & Professor Tanja Sihvonen, University of Vaasa

The technologies and tools that we have come to refer to collectively as “artificial intelligence” in recent years have a communicative user interface: they are controlled and interacted with using natural language. The development of communicative artificial intelligence has been rapid and explosive, and it is clear that in chat-based communication, they are already easily on par with humans – we have come a long way from the Turing test and the generic questions posed by the Eliza bot in the 1960s. At the same time, communication studies has also begun to investigate the specific characteristics of AI-aided communication or communication with AI.

However, the long-term efforts, material resources, and discursive work on which the current “artificial intelligence hype” is based have been overshadowed by leaps and bounds, amazement at new tool versions, and general enthusiasm. Significant research networks and decades of development work can be found behind the scenes. Training artificial intelligence models has been a long-term endeavor, utilizing data accumulated throughout human history. Much of the material used has not been subject to copyright or compensation agreements. AI also has its own stranglehold on the material world: we have only recently started to discuss huge data centers where artificial intelligence applications are run, and whose construction and maintenance require an unimaginable amount of energy, natural resources, and human labor.

Although AI is primarily thought of in terms of cloud services and intangible capital, our working group wishes to emphasize its networked dependencies and material connections. Communicative AI in particular should be evaluated in terms of its functionality in the context of human-machine communication. This draws attention to the roots and apparent rootlessness of AI: its outputs do not originate from anywhere in particular, yet they come from everywhere. The communication capabilities of artificial intelligence can be traced back, on the one hand, to the vast teaching materials that have been and continue to be used to train models and, on the other hand, to computing and the operation of data centers, which consume enormous resources. At the same time, all our understanding of artificial intelligence is constructed and materialized specifically in communication: in interaction with artificial intelligence, but also in the media and corporate discourse about it.

We invite submissions to the working group that examine communication phenomena related to AI in a critical and embodied fashion. The topic can be related to any area of communication studies.

You can submit your proposal in Finnish or English.