Universities invest in a future generation of AI experts

The University of Bergen (UiB) intends to recruit 19 experienced researchers from a broad range of disciplines for advanced interdisciplinary training and career development in artificial intelligence, with a focus on the field’s ethical development.

Starting from 2024, the LEAD AI programme will run for five years and offer joint interdisciplinary training. In addition to funding from UiB faculties of approximately EUR8 million (US$8.78 million), an amount of EUR33 million is expected to come from the European Union’s co-financing scheme, COFUND, which falls under the Marie Sklodowska-Curie Actions in Horizon Europe.

“We need a new generation of AI experts who can exploit the immense possibilities that AI brings to all fields while making AI solutions that are fair, transparent and trustworthy. LEAD AI provides the cross-disciplinary excellence to train these new researchers”, the university said in its Horizon COFUND application.

“Society is undergoing a major transition through digitalisation, with AI taking an increasingly vital role. In scientific research AI enables new research methods and questions, requiring a redesign of education and research training.”

The application noted that while the ability of large language models like ChatGPT and DALL-E to generate high-quality text, speech and pictures had brought the technology to the wider public in 2022-23, “major challenges have been identified, and in many cases, AI has been found to be wrong, biased, unfair, discriminating, and contain hidden propaganda”.

Universities: ‘pivotal’ in AI training

Universities, it said, were “pivotal” in educating and training future experts of AI. “We need experts who can integrate different perspectives and insights ranging from humanities, social sciences, and law to STEM and medicine in order to enable novel, responsible AI solutions.”

UiB Deputy Rector Pinar Heggernes, who also heads the steering committee for UiB AI, said the grant from the EU “underlines the strength of the interdisciplinary expertise in artificial intelligence at UiB.

“We are very pleased that we have now received EU funding to support these postdocs and facilitate joint activities for them across disciplines,” she said.

An article on the university website said LEAD AI at UiB will enable UiB to fully exploit its cross-disciplinary strengths in AI research, providing experienced researchers in fields as diverse as computer science, medicine, law, psychology, information science and the humanities with disciplinary expertise and broader knowledge of AI research needed to develop “trustworthy and ethical AI that benefits society”.

The university said it will recruit both local and international researchers and employ them in the postdoctoral positions for three to four years.

“International mobility is an important element of the programme. UiB will collaborate with research institutions in other countries and with partners in relevant sectors in Western Norway,” it said. The project will also have international collaboration with institutions such as the University of British Columbia, Copenhagen University and Norwegian companies.

Benefits to institution

According to Heggernes, LEAD AI will not only strengthen the competence of postdoctoral fellows; the “entire university” will be strengthened in the field “by inviting young researchers regardless of discipline to the programme's common arenas for research on artificial intelligence”.

The programme would also “promote institutional change and cutting-edge research on AI, while providing researchers with unique research and training opportunities, personalised guidance, and excellent working conditions”, the university said.

The UiB COFUND application received positive evaluations from the European Commission. The feedback from the panel that assessed the application states that the project proposal “well supports Norway's open approach to digitalisation and will deliver employees to the region”.

The feedback also noted that interdisciplinary research options are “well supported by formal and informal networking events, interdisciplinary co-supervision, and opportunities to work with interdisciplinary centres”.

Rector Margareth Hagen said she is pleased that UiB is now becoming even stronger in the field of artificial intelligence: “I am proud of what we have achieved through UiB AI in a short time, thanks to joint enthusiasm and efforts across our skilled AI communities.”

University of Oslo COFUND project

Other universities in Norway will also benefit from the COFUND grant.

According to Morten Dæhlen, centre leader at the dScience Centre for Computational and Data Science at the University of Oslo (UiO), a significant portion of UiO’s COFUND-grant (approximately EUR5 million) will be dedicated to a project known as Data Science Training (DSTrain).

The project, coordinated by the Science centre, will see the employment of 36 researchers in three-year postdoctoral positions in two phases. “In total this will fund 108 annual contracts in computer science at the Faculty of Natural Sciences,” said Dæhlen.

“Our programme with 36 candidates is somewhat broader than artificial intelligence since we are focussing upon ‘data science’ which is the basis for machine learning and artificial intelligence to function. These 36 candidates will come in addition to the 350 doctoral and post-doctoral candidates we already have across UiO.

“I tend to think that Norway, given the population size, is managing rather well within digitalisation. Both scientifically and as a nation we are in front in applying digital solutions. This we have done for a long time, but we might say that we now (with the COFUND projects) are strengthening our position in important areas of digitalisation,” Dæhlen said.

Preconditions for success

He said it was important to communicate that UiO has very good preconditions for success with the additional 36 post-doctoral candidates within data science because over the last 20 to 30 years, the university had built up several leading international research groups in the area.

In addition, the project had included 20 partners from business, the public sector and research institutes. Furthermore, there was “extensive collaboration within a broad network with university partners all over the world, and in this EU project notably with a focus on leading research-intensive universities in Europe”.

Dæhlen said the DSTrain project was a “methodological programme directed towards understanding or solving both scientific and industrial problems where machine learning, artificial intelligence and knowledge representation have a clear focus on green changes, not least because the new AI-solutions through enormous energy consumption have a significant carbon footprint”.

Dæhlen added: “I am looking forward to developing a suitable cooperation with the AI LEAD-programme at UiB, for instance, through seminars.”

Goran Melin, a higher education and research expert at Technopolis Group in Stockholm, told University World News the new initiatives at both UiB and UiO were “positive”. He described the European Commission’s part-funding through COFUND as a “quality mark”.

“It is from my perspective critical that we get more solid research done regarding the consequences through all our society of the rapid AI development.

“These new post-doc programmes are in this respect novel,” he said.

Per Magnus Kommandantvold, who serves as the EU national contact point in Norway for Horizon Europe in the Marie Sklodowska-Curie Actions and the European Research Council, described the developments at the two universities as a “great achievement”.

“The leveraging of competitive MSCA co-funding to an already impressive initiative, makes UiB and UiO significant and visible players in one of the scientific frontiers these days: artificial intelligence. That is good for the universities, Norway and the European scientific community," he told University World News.