AI Sweden connects the dots to keep the country competitive
With world-class research institute in artificial intelligence (AI), Sweden keeps up with all the latest ideas – and sometimes even goes ahead. But blue sky research does not always lead to practical solutions that can be used by industry. That’s where AI Sweden plays a key role.
An important first step in the development of AI Sweden came when Mikael Ljungblom worked as political advisor to the Swedish minister for digital development. When they traveled to see what other countries were doing, Ljungblom and his colleagues saw that countries like Japan and China were investing in AI. Given the importance of technology for competitiveness and societal development, they felt a need to develop an AI center in Sweden. It was 2017.
– I started talking to different players in the ecosystem and I met Mats Nordlund, who is now our colleague, Ljungblom tells Computer Weekly. – He worked in the automotive industry in Gothenburg – and he had the same assignment as me. We locked arms and tried to do this together. We had great support from Vinnovawhich is Sweden’s innovation agency.”
Vinnova provided the basic financing for what would become AI Sweden. Johanna Bergman joined the venture in 2018. She had previously worked with prehospital e-health at Lindholme Science Park in Gothenburg.
– We were given this assignment to set up an AI center in Sweden, says Bergman. “We started by doing a preliminary study – by gathering ideas from all the actors we could think of. We spoke with Swedish industry, with authorities, municipalities, regions, academia and some international actors to try to understand what this should be. It was clear that we would work with applied AI, not basic research. Our mission was to accelerate applied AI in Sweden.
“In February 2019, we inaugurated the center that we called AI Sweden. Partners of all kinds were involved – from startups to companies, academia to the public sector. It was important to spread this from the beginning.”
Ljungblom is now head of communications and public affairs at AI Sweden. Johanna Bergman is now head of strategic investments. She was recently recognized for her achievements by being listed in Forbes 30 Under 30 for science and healthcare in Europe.
“After the inauguration, we immediately started collaborative AI projects – cross-sector and involving different types of organizations – to work on real AI problems,” Bergman told Computer Weekly. “It’s evolved into what we do today. Lindholme Science Park helped us a lot in the beginning and is today one of our host organizations.”
Advantages of a horizontal approach
The fact that AI Sweden operates across sectors differs from most AI research centers in other countries. Instead of taking a vertical approach by focusing on one or two industries, AI Sweden has identified some areas that are important in several sectors.
One such area is edge learning, which is a new paradigm in AI where learning occurs closer to the source of the training data. The models built from the learning process at the edge are then aggregated in a central location where they are shared. The data itself is never transmitted to the central location, which has the desirable side effect of helping to ensure data privacy.
“After the inauguration, we immediately started collaborative AI projects to work on real AI problems. It has evolved into what we do today”
Johanna Bergman, AI Sweden
This differs from traditional machine learning (ML), where the training is centralized and uses a centralized data set. AI Sweden has set up one Edge Learning Labknowing that partners from different sectors – such as the automotive industry, healthcare and the public sector – need to work with the new paradigm.
The Edge Learning Lab includes a testing platform, along with shared resources for developers, data scientists, students, and researchers to learn more about different aspects of edge learning. AI Sweden and partners Zenseact, HPE, Volvo Cars and Ericsson launched the lab in Gothenburg on April 27, 2022. The lab will help different people and organizations to quickly build a work environment for experiments.
Another horizontal focus area is the Swedish language. None of the global technology players bother to develop models for a language like Swedish, so AI Sweden decided to take it up on its own.
– What we are seeing now are major advances in language technology and what we call basic models – huge models that understand natural language and that can generate language, says Bergman. – Almost every week, some organization somewhere in the world releases a new language model – these are large language models that can, for example, generate text from images and images from text. Although we see great progress with English models, not much work is being done for smaller languages like Swedish.”
To help fill this gap, AI Sweden started several projects that address language models for Swedish. The language lab not only helps Swedish companies, but it serves as an example for other cultures whose languages are not yet represented in AI models.
The purpose of the various projects is to develop a national knowledge hub on natural language understanding (NLU) to contribute to stimulating NLU’s work with Swedish. AI Sweden takes up language model training, evaluation and implementation. They also work with ethical and security aspects in several ongoing projects around NLU. Swedish authorities and public organizations are the main stakeholders and will be able to use the models to summarize texts, and for, among other things, translation and writing support.
In addition to cantine learning and Swedish natural language processing, AI Sweden is also working on some other horizontal topics, but has not yet set up programs for them. These include privacy-preserving AI, complex systems, and command and logistics planning. The idea is to find the areas that are relevant to many different sectors.
Networking and collaboration for synergies
– Networks and collaboration are an important part of what we do at AI Sweden, says Ljungblom to Computer Weekly. “We try to connect people who may have common interests without knowing it. Our horizontal approach allows us to come up with completely new – and sometimes surprising – collaborations.”
One of the more surprising synergies is that some hospitals are using models developed by the auto industry for self-driving cars. They use the same models and techniques to predict how the hospital will perform.
Another unexpected collaboration is between the car industry and university researchers who study bird behaviour. The automotive industry uses a data set developed by researchers to categorize the behavior of Baltic sea birds. This dataset is useful for the automotive industry as a proxy dataset for self-driving collaborative projects.
![]()
“Networking and collaboration are an important part of what we do at AI Sweden. We try to connect people who may have common interests without knowing it. Our horizontal approach allows us to come up with completely new – and sometimes surprising – collaborations”
Mikael Ljungblom, AI Sweden
AI Sweden’s approach is to either take what is already happening in pure research and apply it to industry or take what already works in one domain and apply it to another. This is not only beneficial for society, but it also helps Swedish organizations to remain competitive on a global level. By helping Swedish organizations overcome some of the common challenges, AI Sweden accelerates the use of artificial intelligence and helps companies get ahead of the curve in terms of AI adoption.
Talent development and networking
When AI Sweden was founded, it was not obvious that it could play a role in helping to develop the right AI talent in Sweden. But when starting to network with players in different industries, it became clear that AI Sweden was uniquely positioned to discover what talents were needed and to develop consolidated programs that would help develop these skills.
Just last year, AI Sweden started a set of talent programs. These aim not only to develop specific talent, but also to help develop the right company cultures for AI.
AI Sweden also tries to connect the dots and get the different parts of the ecosystem to work together. One of the goals is to ensure that government investments are used in the best possible way. It does this by keeping abreast of who is working on what and finding synergies to get the most out of the funds.
Another way AI Sweden helps accelerate the introduction of AI is by addressing societal challenges. AI is changing the way people work and will probably change industries.
To help prepare organizations for the big changes, AI Sweden is working with partners to ensure they take steps towards becoming more AI-mature. As individual organizations become more mature in applying AI, the entire ecosystem benefits.
– We have found a structured way to do this, says Ljungblom. “We use one AI Maturity Assessment Tool which we adapted from Applied Artificial Intelligence Initiative in Germany. It’s an excellent tool that uses a structured approach to benchmarking improvements and identifying areas to focus on when working with AI. This allows organizations to take the work they do to the next level.”
Sweden’s national advantages
“Finland was quite early in making AI a strategic priority,” says Bergman. “We can gain a lot by cooperating in the Nordics. Several Nordic countries, for example, have very similar languages, so when it comes to language applications there is great potential to work together. It’s starting to happen.”
Ljungblom points out that an interesting thing about Sweden is its decentralized authority structure, with self-governing regions and municipalities. “Therefore, we are forced to find efficient ways to work with many different public entities,” he says. “If we compare the EU with the US and China, we are much more diverse and decentralized, with 27 different member states. In a way, this is reminiscent of the autonomous structure in Sweden, so maybe we have something to offer there.”
Two aspects of Swedish culture give it a competitive advantage in AI. One is that it has always been a technology-friendly country, with a population that is quick to go digital and to use digital tools. It also has a good entrepreneurial system with a good ecosystem of startups. At the same time, there are many very large Swedish companies, such as Ericsson, Saab and Ikea, and these companies tend to keep up with technological developments.
The other advantage is that the Swedish people are used to cooperation. It is part of their culture. In technology, they collaborate across sectors – and between industry and academia. This mindset helps with AI, where an ecosystem is needed and where many of the building blocks can be shared across multiple application domains.
AI Sweden relies on this pervasive attitude of collaboration to help connect the dots.
https://www.youtube.com/watch?v=https://www.youtube.com/watch