GENERAL

Title. Understanding AI (UnAI) (working title, may be revised)
Period: 2023-2025
Core Team: PI: Christoph Schommer. Co-PI: Leon van der Torre. Manager UnAI: Sana Nouzri. Manager AI RoboLab: Igor Tchappi. RF: Ines Crisostomo.
Institution: ILIAS, DCS, FSTM, UNILU. With partners.
Potential institutions: LSC ? Experimental Mathematics Lab (Gabor WIESE) ?
Funding: PSP flagship (FNR). Co-funding: DCS (in kind). Sponsors & donors.
DeadlineSeptember 20.

 

BACKGROUND

Three experiences during the pandemic let us, the computer science department of the university together with partners, to submit this new PSP flagship project:
  1. In 2020 we organized a hybrid course “elements of AI”, which complemented an online course developed in Finland with a seminar given by researchers of our department (co-financed by ministry of state). The participants could interact with us and ask questions.
  2.  During 2022-2024 we are hosting the “Becoming a Computer Scientist” or BeCoS project together with Scienteens in which we teach not only students, but also pupils of high schools (co-financed by FNR via its PSP Flagship program). We stimulate curiosity and receptiveness to computer science, with an emphasis on promoting gender balance in computer science.
  3. In 2022 we are organizing an AI&Art pavilion, in which students and researchers of our department develop interactive exhibitions in collaboration with local artists, and where they meet the general audience and interact with them (co-financed by Esch22).

 

LESSONS LEARNED

From these three activities, we learned several lessons:
  1. There is a large demand for reliable and trustworthy information on AI and related areas like data science and robots (robotics, drones as well as software agents). For example, we got more than a thousand registrations for the elements of AI course, far exceeding our expectations, and we will give it again in 2023 (co-financed again by the ministry of state).
  2. The interest came not only from young people, as we expected, but in particular from their parents! Also groups of VIPs from industry, for example, came to visit us in the AI&Art pavilion with their questions about the role of AI in their industries.
  3. Our visitors sometimes already know a lot about AI, but they don’t know which information they can trust. Interaction is key, online courses are necessary but not sufficient.
  4. Our visitors are not only interested in what AI can do, but also they want to learn and discuss about the dangers of AI. How will it affect their own work, for example. These questions are studied in the interdisciplinary area of AI ethics, and in the social sciences.
  5. The visitors of our AI&Art pavilion like to discuss not only with researchers, but in particular with students in computer science. Maybe because the threshold is lower, maybe because our students know everything about the latest trends.
  6. The exhibitions developed by our students and researchers together with local artists were the key success story of understanding AI by the lay audience. These exhibitions provide a starting point for the discussion between the audience and our students and researchers.
  7. The interaction between artists and researchers and students led in a very natural way also to a critical reflection of the use of AI technology. Maybe because the artists are not trained in technology, or maybe because some of them in their work have a natural critical attitude towards developments in society.

 

OBJECTIVES

These experiences and insights led us to the writing of the Understanding AI project, UnAI in short, where we focus in particular (but not exclusively) on an adult audience,  with the following three objectives (in order of priority):
  1. To increase critical judgement as regards data science, artificial intelligence, and robots,
  2. To provide reliable information to and get feedback from the target audience on these scientific themes, and
  3. To promote life-long learning in data science, AI and robots.

 

METHODOLOGY

Based on our experiences described above, and applying the best practices, we will reach our objectives in the following way:
  1. We maintain the living lab called CCH (for Computational Creativity Hub) created for Esch22 as a meeting place where the general public can interact with students and researchers.
  2. We develop new exhibitions for the living lab CCH in collaboration with social scientists and local artists, addressing in particular a critical reflection on data science, AI and robots.
  3. We organize focussed events addressing specific concerns on data science, AI and robots.

 

OUR INSTITUTION.

Data science, AI and robots are studied not only in the department of computer science, but also in other departments, and in the centers of the university. More generally, digitization is one of the three key themes in the strategic framework of the university, and nowadays data science, AI and robots are applied in nearly every discipline. The living lab is therefore not only a place where students and researchers from computer science meet, but it is also a place to meet students and researchers from other disciplines of the university.

OUR PARTNERS.

  1. FSTM: MATH, RUES, … Data science is a faculty priority and studied also other departments, such as department of mathematics, and the department of engineering.
  2. DP. A critical reflection takes place in the department of philosophy, and AI ethics is one of its key research themes. They are thus an important partner for our first objective.
  3. Art.
  • 3.1. The cultural branch of university (Anouk Wies) for collaboration with local artists and the organization of cultural exhibitions.
  • 3.2. C2DH plays an important role as well, as it hosts many projects now involved in Esch22 projects at the university.
  • 3.3. Partners that can finance artists in residence???
  • 3.4. Partners that connect us to local communities of artists???
  1. Scienteens. The BeCoS project of Scienteens also makes use of the living lab CCH, and we see many synergies between BeCoS and UnAI, such as sharing technicians.
  2. LSC. Together with LSC we developed PSP classic PhotoBooth. Compared to LSC exhibitions, the exhibitions we develop with students and local artists are much more
experimental. The LSC can advise us on the technical development of the exhibitions, just like we provide scientific advice on their exhibitions. Moreover, some of our experimental exhibitions may be further developed into more robust exhibitions by the computer science group in LSC. In addition, we plan to submit joint more focussed projects as well, for example using PSP classic (AI fashion, Scentronix, etc).
  1. LIST. A new group on explainable AI is created in LIST and will be involved in our exhibitions as well, in particular when we receive visitor groups from industry.
  2. University of Seville. Our visiting professor Yolanda Spinola from Spain will develop one of the events (see below).
  3. Ministry of State. Synergy with the hybrid project “elements of AI”, taking place during first three months of 2023 again
Other partners to be discussed.

BUDGET.

The department of computer science is cofunding UnAI by in kind contribution, for example the PIs, managers and administrative support. We ask FNR support to maintain the living lab (100K), student contracts (100K), artists in residence (100K) and events (100K). (Amounts may be revised)

LONG TERM PERSPECTIVE.

Outreach is a key task of our university, and we have strong support from our university management. Additional funding for our activities are expected to come from sponsors and donors. During the pandemic it turned out to be difficult to attract industrial sponsors and donors for the three projects listed in the background section, but we are confident that with the help of Philippe Lamesch, we will be able to change this in the future.

EXAMPLE EVENTS. (To be detailed)

  1. Inclusion. Some parts of society and some parts of the world risk to be excluded by the technological divide. We highlight this issue in collaboration with various partners in Africa, in an event called “I have a dream”. African artists will participate to the artist in residence program. … (Igor)
  2. AI and work. We highlight the impact of AI in all aspects of work by our event on AI fashion. In collaboration with the university of Seville, and as pioneered in the AIFA conference of Esch22, … (Yolanda)
  3. AI education. In this event, in collaboration with the C21 project, SCRIPT and Cisco, we … (Sana)
  4. Women in Computer Science. We have been organizing workshops on gender balance in computer science since 2018. In collaboration with Scienteens, … (Regine, with BeCoS)
  5. AI health. Deep Dreams for medical therapy purposes, or using Computational Creativity for a re-socialisation of psychological / criminal cases (Christoph)
  6. Sustainability and AI. Green IT, the dilemma of power consumption, etc (TBD)
—-
From the documents:
https://www.fnr.lu/funding-instruments/psp-flagship/
“ Projects are expected to achieve qualitative aims like increasing scientific literacy, explaining the scientific method, stimulating curiosity and receptiveness to scientific themes, increasing critical judgement as regards scientific themes, developing an understanding of scientific professions, getting feedback from the target audience on a scientific theme, promoting gender-balance in science, promoting life-long learning, training “vectors” that will enable better communication towards the general public etc. Applicants will have to include an evaluation plan explaining how they foresee to measure the outcomes/impact of their projects.”

 

LINDA REMARKS/ISSUES

The following questions would be important to answer with your project:

  • Why was the new project not directly integrated into the BeCos Flagship project? Why is now the right time to submit this project?
  • What is new and different in this project compared to the part in the BeCos Flagship that was accepted last year?
  • How would you guarantee the sustainability beyond the official duration of both projects, the one in the BeCos project and the one you will submit?
  • It will be critical to show the expert panel that this is a separate project from BeCos with new innovative elements and a different audience to address. As for the FNR rules, nothing speaks against you applying for an additional PSP Flagship, however please beware of double funding, which is not possible, and also to not overbook your own time investment into the project such that all ongoing projects remain feasible.

ANOUK REMARKS/ISSUES

  • To have a strong and convincing concept, it seems to me that it’s important to get the right balance between art and research (AI) and define the role of each of them. Our goal on both sides is to come up with an innovative and qualitative project, where everyone can find his/her interest and professional recognition. The project’s title then also needs to reflect this parity (or be clear about the outcome).
  • It might also be interesting to check with colleagues from the FHSE as they have been treating themes as AI and influences of technology on our society etc.
  • I also see some interfaces with the project “Sound of data”, where data is supposed to be translated in visual designs at some point (next to sound compositions).
  • Regarding partners in the artistic field, it would be great to collaborate with recognized institutions or experts, who might have or would be willing to come up with a budget on their side as well.
  • For the artist part, I was about to ask you if you would be interested in a project I’m looking to organize next year in May/June. Maybe you know the famous French artist ORLAN, quite a phenomenon; she has a robot alter ego and would normally be interested in having it programmed by the University in the framework of an artistic EU event, the European Month of Photography. Please keep this confidential at this stage.
[/vc_column_text][/vc_column][/vc_row]