Nordic State of AI - Central Observations 1/5: Evolution of the Nordic AI landscape

In this five-part blog series, we will dive into the central observations of the Nordic State of AI report, an annual report examining AI adoption across leading Nordic multinational companies by surveying their current use of AI. Overall, companies' adoption of AI has been much aligned with our expectations, but in the big picture, we’re still in the very early stages of realizing AI potential. Big industrial changes are still ahead. 

In this first part of the series, and before diving deeper into the central observations surrounding AI adoption, AI talent, AI infrastructure, and AI readiness, we will take a holistic perspective on the evolution of the Nordic AI landscape as we compare our most recent report with the findings of previous reports.

Comparing report findings over time provides perspective and brings many insights to the surface. The foremost of these is that AI is evolving at such a pace that companies who have not yet embarked on their AI journey risk being left behind and failing to sharpen their competitive edge for the years to come. The urgency to get started is palpable, and unless you started yesterday, the time is now.

The evolving landscape of AI technologies

The number of reported AI technologies in use by surveyed companies has doubled over the span of the past three years. For the first report, there were seven options, whereas for the third report, that number had grown to 14.

This illustrates how different technologies are often grouped together under the mysterious “AI”-label until they become more widely adopted for a particular purpose and become a specific, named technology. The fact that the number of named new technologies has doubled during this short time span is a further indication that prolonging the start of one’s AI journey poses an increased risk of falling behind the competition. 

What has remained the same over the years is that companies report mainly using AI in their products and services. This indicates that companies understand where AI can generate the greatest value. Despite this understanding, however, AI initiatives remain mainly superficial, consisting of scattered PoCs on a rudimentary level that provide little value for a company’s core business and competitive edge in the long run.

The big challenge companies face is moving from an understanding of what should be done to a concrete roadmap for how to actually do it. In our survey, we asked participants about the progress of AI adoption and deployment within companies, as well as how AI development projects are managed.

Only a few companies have a C-level representative championing the implementation and scaling of AI throughout the company. Equally, few have frameworks in place for assessing the success of AI development projects. At the same time, most companies report that they are experimenting with AI and using AI as part of their products, services, and processes. This all points to a sense of urgency, while companies haven’t yet reached a level of maturity where they will fully have digested the strategic role of AI. 

Approaching AI from a strategic perspective requires going beyond the understanding that AI has the greatest potential for value creation when deployed as a core part of products, services, and processes. It includes understanding that all companies are becoming digital at their core as software is eating even traditional industries, and making decisions related to AI infrastructure, ways of working, access to data, internal information flows etc. should be based on that premise. Observing one aspect of this, AI development resources, the first report shows companies relied mainly on internal resources. 

By the time of the second report, this had changed to relying mainly on external resources, whereas by the third report, it had evened out to a balanced combination of both external and internal resources, both in terms of expertise and tools. What this means is that similarly to the number of named AI technologies available, the tools around AI development and infrastructure are also increasing, and companies are starting to understand the infrastructural requirements of AI deployment.

Key challenges in building and scaling AI capabilities

With the rapid technological development of AI comes a surge of new AI products. In addition to tooling for infrastructure and development, everything from personal assistants to image generators and co-pilots are now readily available. This poses new questions for companies. When is it better to buy off-the-shelf AI products, and when should you build custom AI? 

The closer to the core of your business that you are looking to deploy AI, the more sense it makes to build custom AI. There are many arguments to support this, e.g.related to data security, privacy, questions of control and transparency, and not to mention enduring value creation with AI. When using generic AI products, such as productivity tools, there is seldom full transparency in terms of what happens to the data fed to the AI models when using them, nor how and with what type of data they have been trained. When you build and deploy custom AI, you are in control of the data and can train models in a way that is aligned with your business goals, competitive edge, and other factors that might be relevant to you, for instance, your company values, ways of working, and brand.

While many things change over time, some things remain the same. One example of this is that the challenges in scaling the use of AI throughout a company have remained more or less the same over the years. 

The main challenge companies face is the lack of talent, and unfortunately it seems unlikely that the situation will improve anytime soon. Not only is there ample competition for the most advanced technological skill, but there is also a need for increased AI literacy among non-technical employees and domain knowledge among technical employees to enable smooth collaboration and innovation in multidisciplinary teams.

In addition to challenges related to AI talent, other key challenges that have persisted over the years include a lack of shared practices related to data and a lack of clarity in terms of business strategy and the role of AI in it. In our upcoming blog posts, we will explore these topics in more detail, covering AI adoption, infrastructure, talent, and readiness. 

Stay tuned, and in the meanwhile, download our report for more insights.

‍

_{The Nordic State of AI report explores AI adoption across industry-leading Nordic multinational companies, cutting through the hype and providing insights on how to create value with AI. The report gives business leaders, academics, and policymakers a comprehensive overview of the latest changes and developments in Nordic AI. The report provides insights on creating value with AI by understanding what technologies to use, the role of AI infrastructure choices, and when to buy AI “off-the-shelf” versus building custom AI.}