Periodic Reporting for period 1 - TRAINCREASE (From Social Interaction to Abstract Concepts and Words: Towards Human-centered Technology Development)
Periodo di rendicontazione: 2021-01-01 al 2022-03-31
The project aims at i) strengthening the interdisciplinary field of research on the emergence, understanding, and use of abstract concepts and words in human and human-AI interaction and ii) strengthening the research and innovation potential of the University of Warsaw within this field. For the realization of these objectives the University of Warsaw joined forces with 3 European leaders in research on abstraction (Roma La Sapienza), social bases for the emergence of communication (Aarhus University) and Human-AI interactions (University of Manchester).
Understanding that abstract concepts and words arise in interaction and for the purpose of human coordination has a transformative potential for the field of research on abstraction in ERA. Adding the social-interactive dimension engages multiple viewpoints and new methodologies in the field of study: besides cognitive psychology and cognitive sciences, disciplines such as ecological psychology, anthropology, sociology or robotics. This becomes vital in a society, in which smart robotics and artificial intelligence play an increasingly important role in extending human cognition and mediating human interactions and in the times when computational models are trusted with increasingly broad scopes of human activity. The awareness of this social and regulatory character of human knowledge systems, including conceptual knowledge, language, and formal systems will lead to a better understanding that “going computational” in AI-based technology development means efficiency but may also mean distancing from the important first-person human experiences and creating divides along the lines of technical skills.
Traincrease addresses these challenges by stabilizing the transformation of the field of research on abstract cognition from an individual to social dimension on the theoretical, methodological and empirical level. In six work packages the main challenges of the field are identified and methods are developed to study abstraction in embodied human and human-AI interactions. Preliminary ideas are tested within theoretical and empirical “Microtwinnings” (training projects for the University of Warsaw young researchers and staff in collaboration with the Partners). In parallel, Traincrease resources are engaged for introducing transfer of expertise and innovation on the research-management, administrative and educational levels. The aim is to develop a knowledgeable and reliable hub for research on social-interactive genesis and impact of abstract systems, which strengthens the ERA in its striving for inclusive and responsible technological development.
Understanding that abstract concepts and words arise in interaction and for the purpose of human coordination has a transformative potential for the field of research on abstraction in ERA. Adding the social-interactive dimension engages multiple viewpoints and new methodologies in the field of study: besides cognitive psychology and cognitive sciences, disciplines such as ecological psychology, anthropology, sociology or robotics. This becomes vital in a society, in which smart robotics and artificial intelligence play an increasingly important role in extending human cognition and mediating human interactions and in the times when computational models are trusted with increasingly broad scopes of human activity. The awareness of this social and regulatory character of human knowledge systems, including conceptual knowledge, language, and formal systems will lead to a better understanding that “going computational” in AI-based technology development means efficiency but may also mean distancing from the important first-person human experiences and creating divides along the lines of technical skills.
Traincrease addresses these challenges by stabilizing the transformation of the field of research on abstract cognition from an individual to social dimension on the theoretical, methodological and empirical level. In six work packages the main challenges of the field are identified and methods are developed to study abstraction in embodied human and human-AI interactions. Preliminary ideas are tested within theoretical and empirical “Microtwinnings” (training projects for the University of Warsaw young researchers and staff in collaboration with the Partners). In parallel, Traincrease resources are engaged for introducing transfer of expertise and innovation on the research-management, administrative and educational levels. The aim is to develop a knowledgeable and reliable hub for research on social-interactive genesis and impact of abstract systems, which strengthens the ERA in its striving for inclusive and responsible technological development.
In the first part of the project, the excellence in the domain was built on the theoretical level (Work package 1) and on the level of methods integration (Work package 2), while, in parallel, work has been performed to train University of Warsaw ESRs and staff in key skills and processes leading to lasting expertise in research management and funding applications (Work package 3), sustaining the changes on the level of the University of Warsaw infrastructure (Work package 4), and disseminating the results (Work package 5).
WP1: Actions aimed at an overview of the current world expertise on the topic, main challenges and proposals on tackling them:
• A 3-day Symposium “New/Fresh Perspectives on Abstract Concepts: 22 key figures in research on concepts and abstraction delivered open lectures summarizing the state of the field and participated in debates focused on the Traincrease goals.
• Summarizing the challenges (Milestone 1), serving as a roadmap for planned publications and for a call for internal training projects
• Proposal for a special issue in Philosophical Transactions B, one of the highest-impact journals in the field. The Issue “Concepts in interaction: social engagement and inner experiences” is in the final stages of realization, with 22 papers accepted for publication.
• Nine training projects (Microtwinnings) accepted for realization, involving young researchers mostly from the University of Warsaw and experienced researchers at the Partners.
• Special session on abstraction at the international conference of the Scandinavian Association for Language and Cognition, submitted and accepted.
WP2: Actions related to Methods the first reporting period included:
• 1 Workshop in Denmark (Aarhus University) “Cultural evolution in the laboratory”
• Application to IEEE RAS for the Summer School in social robotics hosted by the University of Warsaw
• 3 individual research visits completed
• 2 subsequent Workshops designed in detail
WP3: Actions related to the transfer of knowledge and skills to University of Warsaw:
• 4 lectures and 2 workshops by external experts (4 in-person, 2 online)
• 3 MA and 6 other research projects supervised by the experienced researchers from the Partners’ institutions
• Doctoral Network application, led by Roma La Sapienza.
WP4: Actions towards sustaining the changes at the University of Warsaw:
• Consolidation of the Human Interactivity and Language Lab (HILL) as a research group, with the assigned office space, visual identification and own media channels
• A visit of University of Warsaw research staff to Aarhus University and 5 meetings among Partners and University of Warsaw and Faculty of Psychology administration, detailing the needs
• A new position of Lab Coordinator at the Faculty of Psychology, funded by the University of Warsaw.
WP5: Outreach and dissemination activities:
• Two papers published by the Partners, Special Issue and three further papers accepted
• Presentation of the Traincrease project at the Polish National Contact Point
• Creating and managing the webpage and Twitter, Facebook, and YouTube accounts for Traincrease
• Quarterly Newsletter for the project (5 issues within the reporting period)
• Initiating collaboration with the Copernicus Science Centre, the largest science museum in Poland
WP1: Actions aimed at an overview of the current world expertise on the topic, main challenges and proposals on tackling them:
• A 3-day Symposium “New/Fresh Perspectives on Abstract Concepts: 22 key figures in research on concepts and abstraction delivered open lectures summarizing the state of the field and participated in debates focused on the Traincrease goals.
• Summarizing the challenges (Milestone 1), serving as a roadmap for planned publications and for a call for internal training projects
• Proposal for a special issue in Philosophical Transactions B, one of the highest-impact journals in the field. The Issue “Concepts in interaction: social engagement and inner experiences” is in the final stages of realization, with 22 papers accepted for publication.
• Nine training projects (Microtwinnings) accepted for realization, involving young researchers mostly from the University of Warsaw and experienced researchers at the Partners.
• Special session on abstraction at the international conference of the Scandinavian Association for Language and Cognition, submitted and accepted.
WP2: Actions related to Methods the first reporting period included:
• 1 Workshop in Denmark (Aarhus University) “Cultural evolution in the laboratory”
• Application to IEEE RAS for the Summer School in social robotics hosted by the University of Warsaw
• 3 individual research visits completed
• 2 subsequent Workshops designed in detail
WP3: Actions related to the transfer of knowledge and skills to University of Warsaw:
• 4 lectures and 2 workshops by external experts (4 in-person, 2 online)
• 3 MA and 6 other research projects supervised by the experienced researchers from the Partners’ institutions
• Doctoral Network application, led by Roma La Sapienza.
WP4: Actions towards sustaining the changes at the University of Warsaw:
• Consolidation of the Human Interactivity and Language Lab (HILL) as a research group, with the assigned office space, visual identification and own media channels
• A visit of University of Warsaw research staff to Aarhus University and 5 meetings among Partners and University of Warsaw and Faculty of Psychology administration, detailing the needs
• A new position of Lab Coordinator at the Faculty of Psychology, funded by the University of Warsaw.
WP5: Outreach and dissemination activities:
• Two papers published by the Partners, Special Issue and three further papers accepted
• Presentation of the Traincrease project at the Polish National Contact Point
• Creating and managing the webpage and Twitter, Facebook, and YouTube accounts for Traincrease
• Quarterly Newsletter for the project (5 issues within the reporting period)
• Initiating collaboration with the Copernicus Science Centre, the largest science museum in Poland
The interdisciplinary work performed so far shows that abstractness/concreteness distinction cannot be made along a simple unitary dimension. Theoretical analyses lead to novel conceptual typologies grounded in interaction, history and first-person experience. They point to novel methods to study language-controlled human interaction and language and algorithms-controlled human-AI interaction, which are being developed within Traincrease.
Investigating the details of the process of abstraction will inform the decisions regarding inserting AI-based processes as parts, or stages, of human interaction. Importantly, understanding the nature of abstract concepts and words can take human-machine interaction to new levels both in terms of being able to include abstract relations in the control over artificial systems and, in the opposite direction, making the workings of AI algorithms understandable for human users.
The first 15 months of the project already brought lasting changes for the University of Warsaw. The changes concern knowledge transfer, involvement in key debates in the field on the European and world level, and the consolidation of the team working on the topic. Lasting changes at the level of the Faculty include a novel function of Lab Coordinator. The presence of TRAINCREASE at the Faculty of Psychology resulted in enticing other researchers to apply for the EC projects.
Investigating the details of the process of abstraction will inform the decisions regarding inserting AI-based processes as parts, or stages, of human interaction. Importantly, understanding the nature of abstract concepts and words can take human-machine interaction to new levels both in terms of being able to include abstract relations in the control over artificial systems and, in the opposite direction, making the workings of AI algorithms understandable for human users.
The first 15 months of the project already brought lasting changes for the University of Warsaw. The changes concern knowledge transfer, involvement in key debates in the field on the European and world level, and the consolidation of the team working on the topic. Lasting changes at the level of the Faculty include a novel function of Lab Coordinator. The presence of TRAINCREASE at the Faculty of Psychology resulted in enticing other researchers to apply for the EC projects.