J.
We need to talk about language technologies!

Three problems have converged to create a “perfect storm” for society: 1) some language-related concepts (e.g. translation) are widely misunderstood by people outside the profession; 2) data-driven language processing tools (e.g. Large Language Models, neural machine translation, online concordancers) are no longer in the hands of experts but of everyone; 3) these tools are easy to use uncritically. The result: increasing misuse of language technologies with potentially serious consequences (e.g. in healthcare, legal contexts).

Users need support to improve their AI/language technology literacy, but it’s not always obvious how researchers can help. Increasingly, society needs researchers to engage in science communication to transform complex ideas (e.g. AI, machine translation, corpus linguistics) into accessible guidance that will empower people to use these tools ethically and responsibly. This activity combines the hot topics of AI-based tools and science communication to enable students and researchers to undertake the social responsibility of supporting AI/language technology literacy in society. The goal is to equip these early career researchers with appropriate background knowledge as well as concrete ideas and preliminary toolkits that they can use to engage in science communication activities about their own research as well as more generally on the hot topic of supporting the development of AI/language technology literacy and responsible tool use in their own communities.

The activity will take the form of an interactive workshop where participants are invited to 1) reflect on the need for improved AI/language technology literacy in society, and 2) what role they can play in helping to achieve it. The activity will begin with an introductory lecture (30 minutes) by the workshop facilitator, who will outline the essentials of how AI-based language technologies work, some of the ethical concerns (e.g. Where does the data come from? Is it cheating to use AI tools?), some of the conditions that contribute to misperceptions and misuse of these tools in wider society, as well as some of the consequences. Next, participants will brainstorm some ideas for effective science communication techniques (e.g. metaphors, elevator pitches, infographics) (15 minutes) before dividing into smaller groups to flesh out some concrete ideas for a preliminary toolkit to support AI/language technology literacy (or a related topic based on their own research) in their communities (30 minutes). Finally, the participants will come back together to share the results of the breakout activity (30 minutes), followed by a final summing up of important take-away messages (15 minutes).