Hybrid Intelligence is a 10-year NWO-funded gravitation project (2019-2029) lead by prof. Frank van Harmelen (VU). It is a Dutch national research program formed initially by a collaboration of top AI researchers from the VU Amsterdam, the University of Amsterdam, the TU Delft, and the Universities of Groningen, Leiden, and Utrecht. I was involved with the proposal while I was still working at TUD, and continued my involvement after moving to University of Twente. UT joined as an associate partner in 2021. Involved researchers work in areas such as machine learning, knowledge representation, natural language understanding & generation, information retrieval, multi-agent systems, psychology, multimodal interaction, social robotics, AI & law and ethics of technology.
I serve in various coordinating roles in HI, as well as in supervision of a number of PhD candidates.
Project description
Hybrid Intelligence (HI) is the combination of human and machine intelligence, expanding human intellect instead of replacing it. HI takes human expertise and intentionality into account when making meaningful decisions and perform appropriate actions, together with ethical, legal and societal values. The project’s goal is to design Hybrid Intelligent systems, an approach to Artificial Intelligence that puts humans at the centre, changing the course of the ongoing AI revolution.
Current AI technology surpasses humans in many pattern recognition and machine learning tasks, but it falls short on general world knowledge, common sense, and the human capabilities of (i) Collaboration, (ii) Adaptivity, (iii) Responsibility and (iv) Explainability of norms and values (CARE). These challenges are being addressed in four corresponding research lines. An overview paper describing the vision of the project and the CARE dimensions was published in IEEE Computer (2020).
Coordination: ethics & HI
I serve in several coordinating roles connected with responsibility and ethics in HI. First, I coordinate the Responsible HI research line together with prof. Bart Verheij (RUG). In this research line we investigate how we can align humans and AI systems, accounting for human values, norms, motives, commitments, and goals during system design (“ethics in design”) and at run-time (“ethics by design”). The research line currently comprises 7 PhD projects, addressing topics such as opacity of machine learning systems, value-aligned support agents and ethical aspects of socio-technical system development. We are also shaping a project-wide discussion of what it means to do HI research in a responsible way.
Second, I am co-chair of the HI-Esdit committee together with prof. Ibo van de Poel (TUD). This committee is tasked with stimulating collaboration with another gravitation program, namely Ethics of Socially Disruptive Technologies (Esdit), which was awarded in the same funding round. The two projects have initiated activities to establish connections, since: 1) addressing ethical aspects forms an important part of creating hybrid intelligent systems that truly support people, and 2) Artificial Intelligence is an important socially disruptive technology considered in the Esdit project.
PhD supervision
I am involved with several PhD projects associated with HI on human-AI alignment through interaction. We investigate access to knowledge graphs and user modelling via interaction and conversation, both from a theoretical perspective developing representation & reasoning formalisms, as well as from a user experience perspective.
Specifically, I am daily supervisor of Pei-Yu Chen (started 2021 at TUD) who works on Interactive Machine Reasoning for Responsible Hybrid Intelligence. The supervising team further consists of dr. Myrthe Tielman (TUD), prof. Catholijn M. Jonker (TUD) and prof. Dirk Heylen (UT). Further, I am co-supervisor of Loan Ho (started 2021 at VU) who works on Knowledge Representation Formalisms for Hybrid Intelligence. First supervisors are dr. Stefan Schlobach (VU) and dr. Victor de Boer (VU), and co-supervisor is dr. Myrthe Tielman (TUD). Finally, I have a PhD vacancy on the application of non-monotonic reasoning techniques for user modelling in personal agents. This PhD candidate will be based at the UT and co-supervised with prof. Dirk Heylen (UT) and dr. Shenghui Wang (UT).