Reflections on AI Day 2022
The year’s biggest showcase of AI research in Finland was held for the sixth time on November 16, 2022 at Aalto University.
Photo: Matti Ahlgren/Aalto University
400+ ATTENDEES
58 POSTERS
20 SCIENTIFIC TALKS
8 INDUSTRY TALKS
PANEL: SOLVING GLOBAL CHALLENGES WITH AI
From moving to Mars to engineering the fastest sailboat with reinforcement learning, the panelists covered a lot of ground. The biggest obstacle in deploying artificial intelligence is not technical, but rather ethical and regulatory, they concluded.
Photo: Matti Ahlgren/Aalto University
SCIENTIFIC OR SOCIETAL IMPACT?
Coming from ‘opposite’ camps of industry and academia, Silo AI CEO Peter Sarlin and FCAI director and Aalto University professor Samuel Kaski discussed collaboration, talent, and hype in the AI space.
We have gone through many hype and piloting phases, but now according to the second Nordic State of AI report published by Silo AI, more and more activity in the AI space is moving towards products and R&D that are in the core of companies’ business. That is where the biggest value is being created — Peter Sarlin
“AI is a set of technologies that are ultimately making problem-solving more efficient,” claimed Samuel Kaski. This can accelerate, for instance, science and research, which are analogous to R&D in business sector. At best, we can carry over many of the solutions used in research and science to the R&D side. Companies are already applying AI in production and products, however use of AI for solving problems is more novel.
How to make research–industry collaboration more effective?
Kaski proposed that one possibility is to deploy tri-party collaboration. For instance, there could be a company from the forest industry interested in solving a problem for their customers, a university AI researcher who would participate in solving the related methodological problems, and an expert company that could finalize the solution together with the forest company. There already are a few examples of such collaboration and this would be an impactful way of sharing the know-how between all parties. Fundamental university research in the field of AI is actually not so dissimilar from applied R&D.
Another issue is the bottleneck of human expertise. FCAI, for instance, has a lot of collaboration with companies through sponsorship of doctoral research supervised by FCAI faculty.
Finland is currently doing fairly well with attracting new short-term AI talent to Finland, for example doctoral students and postdoc researchers. “Now Finland needs to be able to keep them by offering an attractive environment as well as interesting enough jobs and challenges to solve,” said Peter Sarlin.