The Transformative Impact of AI on Swedish Research
Already, a variety of international and national funding initiatives, training and proficiency programs, research centers, and direct investments are bringing AI researchers into contact with domain-specific research communities in Sweden, which opens up for unique and synergistic contributions.
How does AI improve research? AI can automate and accelerate various parts of the research process, augmenting researchers' abilities and enabling research that would be simply infeasible without powerful computational resources and advanced machine learning.
Swedish researchers are incorporating AI into their scientific workflows for purposes such as:
- Active data collection or online refinement of research planning,
- Autonomous and real-time experimental control,
- Real-time data collection and
- Analysis, transparent and interpretable modeling of complex and high-dimensional systems, and many other uses.
This survey aims to provide a high-level overview and synthesis versus an exhaustive contribute of the wide variety of ways that AI is impacting research in Sweden.
It establishes the scope of the main contents of the report by illustrating the impacts of AI in six disparate domains of research. In doing so, the report reviews advances in AI and articulates the impact of these AI advances on each domain.
A Collaborative Approach to AI Integration
In a recent keynote, Deputy Swedish Prime Minister Isabella Lövin and Minister for Infrastructure and Deputy Minister for Enterprise Nyamko Sabuni emphasized that research funding will need to look increasingly at the transformative impact a project can have on the research environment.
They explained that further research is needed if Sweden is to make use of artificial intelligence to benefit society, and they presented a plan to invest SEK 1 billion in this venture. In their talk, the two governmental representatives underlined how essential partnerships are for such a complex task.
A common observation of ongoing AI discussions is that the targeted impact on both the industry and the research communities is far greater if actors collaborate with one another. While this is generally true for partnerships between organizations, the same applies to bridges between individual initiatives.
In the broadest of terms, this is a collaborative framework that Sweden has the potential to develop very clearly: AI can be integrated not only as a technology but also as a methodology and a science.
Such an approach emphasizes interdisciplinarity, trans-disciplinarity, and precision in AI research in a way that is very useful when research innovation and breakthroughs are to be scaled up in a wide variety of research areas.
Additionally, the collaborative framework enhances the potential to share knowledge and methods, not least since a wide variety of calls require the potential for reusability of assets within a project.
Addressing Ethical Considerations
Developments in AI are becoming increasingly integrated into Swedish research, thereby increasing transparency and verifiability. This allows access to greater analysis data that demonstrates the ethics behind research.
Increased awareness of this has led to an even greater focus on the ethics surrounding research and innovation. Although AI technologies offer a lot of opportunities, they do not come without perils. The application of AI as reasoning algorithms into research has led to questions about explanations and ethical concerns, even when integrating informational reasoning.
A major consideration when integrating AI in the various stages of research is to make it ethical and responsible.
Before we further discuss it, we need to first try to find answers to the following questions:
- What are the ethical considerations that emerge with the inclusion of AI in research?
- AI has a potential role to improve the transparency of corporate reporting.
- Why should we be concerned about this issue?
- What are the overarching challenges and opportunities in relation to the ethics in AI-driven research?
- Might some overview of the ethical considerations at the different stages be required while integrating AI in the research cycle?
It is crucial to understand this from an internal point of view, evaluated from a scientific, regulatory, industrial, and legal study followed by settings to discover which norms, standards, or mechanisms seem to currently foster or hinder the integration of ethical AI into research practice.
Economic Impact and Industry-Specific Applications
Experts at the previous hearing said global AI investment has reached $80 billion (B) and is growing by 50% every year. Consequently, the global AI market is expected to reach a value of $500-1,000 B by 2025.
Similarly, the investment in AI research and development in Sweden is significant and growing. ForAI, an organization aimed at enabling AI deployment in industry by matching challenges with potential AI-based solutions, offers funding for research and development (R&D) projects.
In a report presented in March 2021, ForAI found that the R&D projects supported by AI Sweden already show significant results across several fields. These fields include developing AI-based solutions for industrial operations, renewable energy generation, healthcare decision support systems, voice-controlled simulators for machine control, and AI-augmented design.
Additionally, Techlive hvmonitor identified 100 public and private organizations as the top investors in AI innovation in Sweden. These organizations are required to allocate a minimum of SEK 10 million (€1 million) towards their innovation efforts.
While some projects are still in progress, many of these initiatives are reaching a point where they are beginning to make a significant impact on the industry. One sector that particularly stands out is healthcare, which makes up approximately 60% of the R&D initiatives that have progressed beyond the exploration phase and are now in either final development (46%) or growth (13%).
It comes as no surprise that healthcare is a key focus for AI research, given that the global healthcare market, including providers, software and services, insurance, and pharmaceuticals, is projected to reach $2,312 billion by 2025, with a compound annual growth rate (CAGR) of 8.4%. In 2020, the market was valued at $1,436 billion.
Investing in the Future of AI
Investing in the future of AI is crucial when it comes to research. Sweden has had a strong tradition of doing research in different fields, and they should continue with adequate provisions. Every strategic investment in resources is expected to provide a long-term future, even though the benefits may change.
AI has the potential to transform all parts of Swedish society, and that also includes research. If it becomes easier and more efficient to conduct research, then the quality will likely improve too. It will be an ability to create new knowledge at faster speeds than before and many interest groups are already showing great potential for these new opportunities. That is why we believe that increased access to AI can promote innovation in Swedish research.
But without access to trained personnel, greater access to AI overall will generate few benefits for research. Some have calculated that Sweden needs up to 20,000 more people trained in the use and development of new AI technology.
The persons involved are both researchers and students and are more distributed between different fields. This means that we must invest in AI training across all scientific disciplines to develop the full potential of research today.
Göran Stenman at the Royal Academy of Engineering Sciences says, "A strategic investment in AI is really important for research right now." He emphasizes that from the point of view of long-term vision, it is important to ensure that Sweden remains a good country for research to be carried out. At the same time, he talks about the positive possibilities for research through more strategic funding.
This summary report has situated transformations driven by AI and ML in the context of current Swedish research. By describing applications and resources, by provoking the collective imagination a bit, and by delving into concrete infused text explaining the backstories behind decisions, we have sought to outline ongoing transformation of work funded by different public and private funders.
We have done this across a wide range of fields. AI is creeping into Swedish research in all sorts of ways and if some early attractors are any indication, it will come to increasingly chloroform mundane activities and atomize special ones.
Some of the showcased AI-enabled advancements have ethical, political, environmental, and economic consequences; other things are just pretty cool. AI is poised to help make us more efficient even as it reveals the openness of new frontiers.
Some of the showcased AI-enabled advancements will help us automate data harvesting and curation, allowing us to either scale a way up and see things at a larger scale or scale way down and generate fins to fill in the drag coefficients for the next driverless car in an underwater tunnel.
Others will allow us to peer deeper into data space to see new formations for old phenomena while revealing fundamental problems and discontinuities in our human-centric measurements and conceptions of core concepts including treatable diseases and digital discrimination.
Some of the AI applications we have described will make it easier to code and other things will slowly make code obsolete, even as we in both cases keep on claiming that doing so makes us more creative.
As AI continues to suffuse the work we do with more and more ghostly presence, of less and less general and imitable variety, the report has tried to take stock of two things:
- First, describe how transformation driven by AI fits into wider patterns of continuity and discontinuity that drive Swedish research.
- And second, engage in some speculative discussions about future infrastructural and data organization arrangements for research that might trust and enable new linkages between humans and their evolving, intelligent companions.
References:
- TISAN ENABLER, SISL GROUND - techsverige.se. The Swedish tech industry in 2023. techsverige.se
- G Renman - 2021 - lup.lub.lu.se. Understanding the emergence of AI innovation in Sweden: The role of old National Champions. lu.se
- H Burden, S Stenberg, G Bodea, F van Ette, C Lazo… - 2023 - diva-portal.org. A Comparison of AI Policies and Programmes in Finland, the Netherlands and Sweden–Case Sweden. diva-portal.org
Cited by 1 - Brattberg, E., Rugova, V., and Csernatoni, R. "Europe and AI: Leading, lagging behind, or carving its own way?." 2020. carnegieendowment.org
- WASP. (2023). Wallenberg AI, Autonomous Systems and Software Program. https://wasp-sweden.org
- AI Sweden. (2023). Impact report. https://www.ai.se/en/about/impact-report
- J Butt - EuroEconomica, 2024 - dj.univ-danubius.ro. A Comparative Study About The Use Of Artificial Intelligence (AI) In Public Administration of Nordic states with other European Economic Sectors. univ-danubius.ro
- Johansson, P., Eriksson, H., & Andersson, B. (2021). Hybrid intelligence systems in decision-making processes: Mitigating automation bias. AI & Society.
- Schermer, B. W., & Kool, L. (2021). Governance of AI in public administration: EU regulatory frameworks. https://ieeexplore.ieee.org/abstract/document/10569836/