Notable on the fresh FAIA State of AI in Finland report is the attitude on focusing on the concrete cases companies are doing in the field AI. Described cases enable much more accessible and richer conversations about the development and capabilities of the AI ecosystem.

These are quick notes and highlights from the perspective of a machine learning startup born in Helsinki that is building concrete cases to demonstrate that we can warm up the AI winter. The period of reduced funding and interest in AI research, due to the lack of evidence of the benefits of AI after long and loud hype.

You can download and read the full report here.

The pudding

You do not need to test all the attractive AI opportunities by yourself necessarily. Use your ecosystem to get proof of feasibility, which is tested by someone else. Focus on getting to know the concrete and easy to understand use cases. And then select the most impactful ones to your portfolio.

Check out the #1 from the Best Cases list in the report. The quote is not mine despite my name being there, but Posti’s head of intelligent automation Jani Rahja’s. Aito boosted the logistic giant’s invoice automation by adding machine learning to churn invoices.

#1 use case out of the top 20 best cases in 2020 State of AI in Finland. Image source: FAIA

Keep on scanning the use cases of this report and the ecosystems near you. I am happy to see that for example, the RPA’s marriage with machine learning is more and more visible in these reports.

Lack of understanding

What shines through the report is (still) a distortion between expectations and realistic results with AI, along with the widespread lack of reliable and usable data.

We did a short very unscientific study about a year ago asking tech people “if you would have an extra ML engineer, what would you do?” We were expecting a wild list of use cases, but mostly heard answers along the lines of “I would ask what on earth would I do with ML that would bring benefits for my team”.

AI ecosystem is booming

We are not alone; there are now 400 AI startups in Finland, with 50 new ones founded every year.

Prerequisites for Finland on being the breeding ground for global AI success are there. Finland is ranked high on AI readiness. Even though Helsinki ranks high, the ecosystem has diversified in terms of geographical spread. Oulu, Turku, Tampere and Pori come to my mind with good examples of exceptional talent and good local scene.

In three years, the number of companies applying AI has almost tripled, yet only 3% of the Finnish companies (employing five or more) utilise AI daily. And of course, we are a geeky nation, full of talented engineers. Majority of the companies who use AI applications (1250) have developed their own tech (750).

COVID impact still mostly invisible

For some, the pandemic environment is an opportunity; for some, this pushes them to think even harder where to focus. And some, unfortunately, will suffer due to their stage of development or securing the cash to survive. Top AI firms have experienced setbacks and productivity bumps when working remotely.

Still, close to 90% of the companies aim to recruit more, with plans of more diversity by hiring foreign nationals.

Is a long winter coming?

According to the report, the situation in Finnish AI investments has not been following (at least yet) the same level of an impact as the hit has been to the overall economy. We stay tuned and follow the situation in our dear AI ecosystem. FAIA will publish the second part of the report in November.

We say let the winter come! After all, this is Finland, and we have long winters every year. We have used to fight back relentlessly when we see that winter is coming. Even if it is AI winter.

Photo by LMM