CEO of NORA – Norwegian Artificial Intelligence Research Consortium
Tell us about your first encounter with artificial intelligence (AI)?
What really opened my eyes was AlexNet, the convolutional neural network (CNN), designed by Alex Krizhevsky and published with Ilya Sutskever and Geoffrey Hinton. AlexNet won the ImageNet Large Scale Visual Recognition Challenge with good margins in 2012, and made me consider switching field from computational neuroscience to artificial intelligence, two fields that in many aspects are related.
What is your competence within the field of AI?
Biology-inspired AI and deep learning.
Why did you develop an interest in AI?
I have always been fascinated by how the brain works, memory, intelligence and perception. Around 2012 I realized that we live in a time when we may develop the theories and computational power to be able to artificially create strong perceptive systems and intelligence. It is simply too exciting not to be part of that.
Can you recommend a relevant book or film about AI?
Related to the talk I am giving (Artificial Intelligence in health care), I will absolutely recommend Eric Topol’s Deep Medicine: How Artificial Intelligence Can Make Healthcare Human Again. One of my favorite movies is Her, a Spike Jonze movie from 2013 about a man who develops a relationship with Samantha, an artificially intelligent virtual assistant personified through a female voice.
Why should we, or should we not, be afraid of AI?
We should not be afraid of AI. We have the skills to develop this tool wisely. However, we should be very aware of its possible negative use, challenges to privacy, automatic warfare etc.
Which field, in your opinion, has the most to benefit from AI – and why?
Today I will answer health care. Recent developments in radiology, pathology, genetics, sensor technology and a number of other areas makes it possible to collect enormous amounts of patient data, data that can provide more reliable diagnoses and can be the foundation for tailored treatments.
How should the use of AI develop in the future?
The first step is to develop good privacy preserving systems, good methods for training AI systems without challenging privacy. In health care, and in society at large, AI systems will be used whenever the AI system has an advantage. This is important for personalization. When it comes to health care personalization and data will be important both for diagnosis and for treatments.
Why should participants tune in during AI+?
You should tune in to get a glimpse of the speed within the field of AI and how this field is about transform the way we do health care.
Klas Pettersen is a speaker at the AI+ virtual conference 2020.