-    Come October 2021, and guess what? It is clear that the real-world complexities – or better frictions – of getting the AI technology to work and be accepted as safe and reliable, will delay its arrival. 

Ieva Martinkeinate said she still drove her own car although an American government secretary in 2016 said there would be selfdriving cars everywhere in 2021.
She cannot order a robot taxi at Gardemoen that could drive her home although Elon Musk in 2019 said there would be a million robot taxis the next year. And the virtual assistants still have å long way to go to be really helpfuI in all situastions.  

-    On top of that, many of us wait for ML to solve complex real-world problems and move beyond toy problems, such as beating the best human champion in the game Go or chess.

-    What technology evolution theory teaches us, is that hyping early stage development of tech is very normal. However, every cycle of excitement leads to pragmatism and hard work, Ieva Martinkeinate said. 

Serious observers, like the The Economist, declared back in 2020 that “AI autumn breeze” is coming with less research activity and attention towards AI back.
Ieva Martinkenaite, however, stated that there is progress in AI and more specifically in machine learning ML. Progress is everywhere as software is helping us to navigate in streets and avoid traffic jams, filtering email spam, recommending our most favorite songs, food or movies, or even reducing power consumption on energy grids. 

Even so, after giving a detailed walk through of specific challenges in the crossroad between academic research and business developers, she asked: is AI still cool?

Her answer:

-    Definitely yes!

-    Because AI hype is diminishing and moves from exuberance to pragmatism. Obviously mostly cited moonshot projects, such as self-driving cars, will be delayed. 

-    Instead, focus is shifting towards AI fighting against the pandemic, climate change and optimizing everything also the mission critical infrastructure. Due to powerful AI hardware and more accessible software, 2022 will see AI solving practical problems and becoming increasingly embedded in business processes, she said.  

-    The socalled weak AI requires hard work and fundamental change in how companies learn – building data pipelines, developing talent, creating software, embracing diversity and responsibility in how we build and use AI, she added. 

Ieva Martinkenaite is chair of NorwAI’s innovation Board. She set high goals for the Research Center: 

-    AI brings breakthrough to research and innovations, which gives hope that the biggest Research Centre on AI in Norway will succeed in spinning out some unique, global unicorns and help our large businesses to optimize processes.

Her hope for the pathway forward was high, as she said research nerds from academia and innovators from business do cross seamlessly for the best – green and sustainable – future of Norway AI. 

-    My experience shows that combining AI research and innovation is not trivial. There is hard work, investments and ambition needed, Ieava Martinkeinate added.