The European Laboratory for Learning and Intelligent Systems (ELLIS) is establishing a project to grow research institutes that focus on the societal impact of AI. Today the nonprofit announced 17 cities in 10 European countries and Israel that it has selected to participate in the project.

Each selected site — in places like Amsterdam, Copenhagen, Helsinki, Tel Aviv, and Zurich — will begin with about half a dozen AI researchers and sustained funding of at least €1.5 million ($1.66 million) a year for the next five years. In an industry known for aggressively hiring top talent, the funding is meant to provide researchers with salaries comparable to those offered by tech giants. Units begin operations in spring 2020.

“A step toward realizing this vision of having these institutes is creating these units, which are a little bit like the seed. [I]f the units are successful, the plan is that it will grow and … eventually become an institute,” AI researcher and ELLIS board member Nuria Oliver told VentureBeat in an interview today at the Neural Information Processing Systems (NeurIPS) conference in Vancouver, Canada.

At NeurIPS in Montreal last year, top machine learning minds in Europe joined together to create ELLIS, which they hope grows into a leading open science organization with labs around the world.

ELLIS was inspired by the Canadian Institute For Advanced Research (CIFAR) initiative to invest in the Canadian AI ecosystem.

At NeurIPS in Vancouver this year, ELLIS also signed a letter of intent with CIFAR’s program on Learning in Machines and Brains, which is codirected by Turing Award winners and deep learning pioneers Yann LeCun and Yoshua Bengio. The organizations plan to take joint actions, such as establishing summer schools and workshops and streamlining review processes for fellowship programs and exchange visits between Canada and Europe.

The goal is to attract and retain top European AI talent and increase the impact of AI research by European authors.

Concern about Europe falling behind the United States and China in AI development was another motivation behind ELLIS, Oliver said. For example, Elsevier finds that Europe writes more AI research papers than any other region on Earth, but EU researchers’ work isn’t as heavily cited as that of contributers elsewhere in the world.

“The reality is that Europe is lagging behind China and U.S. in [research] impact,” Oliver said.

Aside from the need to keep salaries competitive with companies like Google, Facebook, and Microsoft, ELLIS identified teaching load as a barrier for top scientists, who must teach each week instead of advancing their research. ELLIS cited rigid European attitudes about enabling scientists to engage with industry or create their own company as another hindrance.

“Because we want both the scientific excellence and the societal impact — and a big way to get societal impact is through companies or through startups — another criteria is that the people in their units will have the flexibility to engage with industry or spin off their own company while still being an active researcher,” she said.


The audio problem: Learn how new cloud-based API solutions are solving imperfect, frustrating audio in video conferences. Access here