Norway may take a world-leading AI role 

Where is Norwegian AI research heading (where can we expect results short term/long term)?

I can answer this question from the perspective of information access and decision support systems. Many problems in this space are still either too complex or too important to be fully automated with humans removed entirely from the loop (consider, for example, financial decision making or content creation). That said, there are numerous smaller, specific steps within these more complex processes where AI can help assist humans—think of search engines helping with query formulations, forms suggesting automatic categorization of content, or conversational agents being able to tackle simple tasks. Thanks to improved natural language understanding, we can expect to see a lot more of these in the short term and a significant expansion both in terms of scope and complexity in the long term.

Therefore, I can see Norwegian AI researchtaking a world-leading role in these areas. 

How can research facilitate innovation in your opinion?

I can think of opportunities on at least three fronts.

First, there is often a critical performance threshold for methods or components in a given application setting. Research is continuously trying to improve performance to reach (and push beyond) these thresholds. Also, there is a great deal of research effort directed towards ensuring that the “right things” are measured and in a sound way.

Second, the cost of deployment or data annotation can be a barrier to innovation. Research can help lower these barriers by making techniques more accessible (e.g., by building on top of widely used software frameworks and open-sourcing solutions) and by developing methods that are less data-hungry.

Third;

These can lead to innovative products or user experiences.

Looking forward, what is success for NorwAI?

Success comes in many forms. For academic partners, success may be impactful publications or methods/ideas making it to being deployed and used in products. It would also be success to see some of the PhD candidates funded by NorwAI ending up as future leaders in the industry. For industry partners, success may be increase in key performance indicators, the launch of innovative services, or increase in in-house AI expertise. For all project partners, continued and extended collaboration beyond NorwAI would be a sign of success.

Ultimately, success for NorwAI, in addition to all of the above, may be recognition, both within and outside Norway, as an exemplary model of academic-industry collaboration.