Project: Value-Aware Artificial Intelligence
2022-10-01 – 2026-09-30
- Abstract
Value-aware AI means an AI performing human moral functions such as acquiring and maintaining a value system and deciding morally acceptable actions, understanding their intent and properly engaging with them. The EIC-funded VALAWAI project will develop a toolbox to build value-aware AI on two science-based pillars that include an architecture for consciousness and a foundational framework for moral decision-making. The first pillar is inspired by the global neuronal workspace model developed on neurophysiological evidence and psychological data. The second relies on psychology, social cognition and social brain science. The project will focus on social media, which include several negative effects such as disinformation and polarisation, social robots, which can enable manipulation, and medical protocols.
-
Vision language models as values detectors
(2025) VALUE ENGINEERING IN ARTIFICIAL INTELLIGENCE, VALE 2024. In Lecture Notes in Computer Science 15356. p.76-86 -
- Conference Paper
- P1
- open access
'Can you be my mum?' : manipulating social robots in the large language models era
-
- Conference Paper
- P1
- open access
I was blind but now I see : implementing vision-enabled dialogue in social robots
-
- Journal Article
- A1
- open access
Integrating visual context into language models for situated social conversation starters
-
Child speech recognition in human-robot interaction : problem solved?
(2025) SOCIAL ROBOTICS, ICSR + AI 2024, PT II. In Lecture notes in computer science 15562. p.476-486 -
- Conference Paper
- P1
- open access
Values in social robots : implementing inclusive, value-aware human-robot interactions
-
- Journal Article
- A1
- open access
Tactile interaction with social robots influences attitudes and behaviour
-
- Conference Paper
- P1
- open access
Social value alignment in large language models
(2024) VALUE ENGINEERING IN ARTIFICIAL INTELLIGENCE, VALE 2023. In Lecture Notes in Computer Science 14520. p.83-97 -
- Conference Paper
- C1
- open access
AwarePrompt : using diffusion models to create methods for measuring value-aware AI architectures
-
- Conference Paper
- C1
- open access
Towards a Definition of Awareness for Embodied AI