Periodic Reporting for period 2 - TAILOR (Foundations of Trustworthy AI - Integrating Reasoning, Learning and Optimization)
Período documentado: 2022-03-01 hasta 2023-08-31
The purpose of the EU Project TAILOR is to build the capacity to provide the scientific foundations for Trustworthy AI in Europe by developing a network of research excellence centres leveraging and combining learning, optimisation, and reasoning. These systems are meant to provide descriptive, predictive, and prescriptive systems integrating data-driven and knowledge-based approaches.
Artificial Intelligence (AI) has grown in the last ten years at an unprecedented pace. It has been applied to many industrial and service sectors, becoming ubiquitous in our everyday life. More and more often, AI systems are used to suggest decisions to human experts, to propose actions, and to provide predictions. Because these systems might influence our life and have a significant impact on the way we decide, they need to be trustworthy. How can a radiologist trust an AI system analysing medical images? How can a financial broker trust an AI system providing stock price predictions? How can a passenger trust a self-driving car?
These are fundamental questions that require deep analysis and fundamental research activity as well as a new generation of AI talents who are skilled in the scientific foundations of Trustworthy AI, who know how to assess, and how to design, trustworthy AI systems. Some of the current issues related to lack of trust in AI systems are a direct consequence of the massive use of black-box methods relying only on data. We need to define the foundations of a new generation of AI systems not only relying on data-driven approaches, but also on the whole set of AI techniques, including symbolic AI methods, optimization, reasoning, and planning.
Artificial Intelligence (AI) has grown in the last ten years at an unprecedented pace. It has been applied to many industrial and service sectors, becoming ubiquitous in our everyday life. More and more often, AI systems are used to suggest decisions to human experts, to propose actions, and to provide predictions. Because these systems might influence our life and have a significant impact on the way we decide, they need to be trustworthy. How can a radiologist trust an AI system analysing medical images? How can a financial broker trust an AI system providing stock price predictions? How can a passenger trust a self-driving car?
These are fundamental questions that require deep analysis and fundamental research activity as well as a new generation of AI talents who are skilled in the scientific foundations of Trustworthy AI, who know how to assess, and how to design, trustworthy AI systems. Some of the current issues related to lack of trust in AI systems are a direct consequence of the massive use of black-box methods relying only on data. We need to define the foundations of a new generation of AI systems not only relying on data-driven approaches, but also on the whole set of AI techniques, including symbolic AI methods, optimization, reasoning, and planning.
TAILOR has 54 partners from 20 European countries. In addition, around 100 other labs have chosen to be part of the network. Together we assemble 150 leading European research labs and aim to give rise to a culture of sustained collaborations and transformative innovations. The network enables faster research progress by providing a common resource for Europe and the world with easy access to state-of-the-art knowledge and expertise in the foundations of Trustworthy AI.
At the project start, in September 2020, and through the first 18 months, the covid pandemic prevented people from travelling. Networking has relied on digital tools; no physical project meetings were arranged during the period. Despite this, partners have succeeded in collaborating, both on reports, on scientific articles, and on network events. The project had produced >100 scientific articles at M18, of which > 50 were A-level and >14 included two or more partners. TAILOR (co-)organised at least 5 conferences as well as numerous workshops.
Network collaboration is at the heart of TAILOR, and efforts are made to promote and support interaction and collaboration between network members and project partners through tools, development of training material, organisation of training events, PhD and staff exchanges, etc. The overall goals are to foster excellence through shared knowledge and vision, to create a sustainable research capacity in Trustworthy AI which has critical mass and can effectively interact with European funding mechanisms, and to establish a healthy pipeline of future talent. A Connectivity Fund is available to all partners and network members, for funding of collaborative projects and extended research visits. The fund is expected to be even more attractive in the next period, provided that the pandemic does not resurge.
The TAILOR network includes many of Europe’s top AI labs, and we want to reach out to the many other excellent labs and organisations across Europe to work together and create new breakthroughs in AI.
We especially aim to support young researchers to gain valuable experience and nurture the next generation of AI researchers.
Three Theme Development Workshops (TDW) were co-arranged with HumanE AI Net, VISION and CLAIRE. All were set up to bring together key players from specific industry sectors with key AI researchers, as well as other stakeholders. The objectives are to jointly identify strategic AI research areas and challenges, and to initiate further activities to address them. The first TDW was on “AI in the Public Sector”, the second on “Future mobility”, and the third on “AI for Future Healthcare”. All were successful, providing platforms for such discussions that are key in unlocking the full potential of AI in Europe.
The TAILOR Strategic Research and Innovation Roadmap (SRIR) aims to boost research on Trustworthy AI by defining major research challenges. It is focused on the integration of learning, optimization, and reasoning to improve the trustworthiness of AI systems. Academic contributions were fetched from the scientific Work Packages. Applied research directions also came out of the TDWs.
The four ICT-48 networks (AI4Media, ELISE, HumanE-AI NET, TAILOR) and the VISION consortium have founded a joint instrument to support a world-class AI PhD level education programme; the international AI Doctoral Academy (AIDA). AIDA offers access to knowledge and expertise and supports attracting PhD talents in Europe.
At the project start, in September 2020, and through the first 18 months, the covid pandemic prevented people from travelling. Networking has relied on digital tools; no physical project meetings were arranged during the period. Despite this, partners have succeeded in collaborating, both on reports, on scientific articles, and on network events. The project had produced >100 scientific articles at M18, of which > 50 were A-level and >14 included two or more partners. TAILOR (co-)organised at least 5 conferences as well as numerous workshops.
Network collaboration is at the heart of TAILOR, and efforts are made to promote and support interaction and collaboration between network members and project partners through tools, development of training material, organisation of training events, PhD and staff exchanges, etc. The overall goals are to foster excellence through shared knowledge and vision, to create a sustainable research capacity in Trustworthy AI which has critical mass and can effectively interact with European funding mechanisms, and to establish a healthy pipeline of future talent. A Connectivity Fund is available to all partners and network members, for funding of collaborative projects and extended research visits. The fund is expected to be even more attractive in the next period, provided that the pandemic does not resurge.
The TAILOR network includes many of Europe’s top AI labs, and we want to reach out to the many other excellent labs and organisations across Europe to work together and create new breakthroughs in AI.
We especially aim to support young researchers to gain valuable experience and nurture the next generation of AI researchers.
Three Theme Development Workshops (TDW) were co-arranged with HumanE AI Net, VISION and CLAIRE. All were set up to bring together key players from specific industry sectors with key AI researchers, as well as other stakeholders. The objectives are to jointly identify strategic AI research areas and challenges, and to initiate further activities to address them. The first TDW was on “AI in the Public Sector”, the second on “Future mobility”, and the third on “AI for Future Healthcare”. All were successful, providing platforms for such discussions that are key in unlocking the full potential of AI in Europe.
The TAILOR Strategic Research and Innovation Roadmap (SRIR) aims to boost research on Trustworthy AI by defining major research challenges. It is focused on the integration of learning, optimization, and reasoning to improve the trustworthiness of AI systems. Academic contributions were fetched from the scientific Work Packages. Applied research directions also came out of the TDWs.
The four ICT-48 networks (AI4Media, ELISE, HumanE-AI NET, TAILOR) and the VISION consortium have founded a joint instrument to support a world-class AI PhD level education programme; the international AI Doctoral Academy (AIDA). AIDA offers access to knowledge and expertise and supports attracting PhD talents in Europe.
By developing the scientific foundations of Trustworthy AI, major impact both in Europe and globally is expected. Most importantly through the deep and broad basic research performed in TAILOR. We also expect to create new, and grow existing, scientific collaborations both within TAILOR and with other research centres.
The TDWs are expected to identify important research challenges for European industry. These will be important contributions to the SRIR, which will contribute to the strategic research and innovation agenda for AI in Europe, proposing innovative research directions toward Trustworthy AI-based systems, in particular regarding hybridizations between symbolic and neural AI, or between AI and other scientific disciplines.
More Theme Development Workshops will be organised which will further enrich the SRIR – while also, though more indirectly, impacting its evolution by unveiling and highlighting research directions that need to be explored, in the virtuous cycle tightening the connection between Research and Innovation.
TAILOR is a key member of AIDA and supports its goal of coordination of PhD and postdoc educational and training activities on AI, aiming high to become a world-level reference for AI education.
AI and all the key digital technologies that are subsumed by the term AI today are an essential part of the answers to many of the daunting challenges that we are facing. AI will impact the everyday lives of citizens as well as all business sectors. Europe aims for human centred Trustworthy AI, and TAILOR is one important tool to get there. Trustworthiness can only be achieved through the integration of learning, optimisation and reasoning, as neither approach will be sufficient on its own. Thus, TAILOR contributes to a strong academic-public-industrial research network in Europe, with the capacity of providing the scientific basis for AI systems that will be reliable, safe, transparent and respectful of human agency and expectations.
The TDWs are expected to identify important research challenges for European industry. These will be important contributions to the SRIR, which will contribute to the strategic research and innovation agenda for AI in Europe, proposing innovative research directions toward Trustworthy AI-based systems, in particular regarding hybridizations between symbolic and neural AI, or between AI and other scientific disciplines.
More Theme Development Workshops will be organised which will further enrich the SRIR – while also, though more indirectly, impacting its evolution by unveiling and highlighting research directions that need to be explored, in the virtuous cycle tightening the connection between Research and Innovation.
TAILOR is a key member of AIDA and supports its goal of coordination of PhD and postdoc educational and training activities on AI, aiming high to become a world-level reference for AI education.
AI and all the key digital technologies that are subsumed by the term AI today are an essential part of the answers to many of the daunting challenges that we are facing. AI will impact the everyday lives of citizens as well as all business sectors. Europe aims for human centred Trustworthy AI, and TAILOR is one important tool to get there. Trustworthiness can only be achieved through the integration of learning, optimisation and reasoning, as neither approach will be sufficient on its own. Thus, TAILOR contributes to a strong academic-public-industrial research network in Europe, with the capacity of providing the scientific basis for AI systems that will be reliable, safe, transparent and respectful of human agency and expectations.