Welcome to the latest in a series of brief interviews with guest experts from KnowledgeBrief’s Innovation Programme, providing a window into the experts’ latest ideas and new advice for executives.
Following the Innovation Day in September, Dr Viktor Dörfler (VD) from University of Strathclyde Business School gave us an interview to discuss Artificial Intelligence (AI).
KB: What’s the key business challenge that organisations need to address, that your research tackles?
VD: The main challenge for most businesses is to figure out the possible role(s) of Artificial Intelligence (AI) in their particular context. The bottleneck today is not so much in producing further or better AI solutions. We already have excellent AI, one of the best solutions is here in the UK, it is DeepMind. What we don’t have is people who can figure out what to use AI for and how. It is not easy to understand what AI really is, what it can do and how it can be used. The projections about what AI will do may seem wild, but they are pretty consistent: AI is expected to contribute to the UK economy in the range of hundreds of billions of GBP over the next 10-15 years. What each individual business needs to figure out is: what does this mean for me?
KB: What advice would you give to executives, based on your findings?
VD: My advice is based on the inevitability principle: AI is becoming ubiquitous. Therefore, executives need to educate themselves about AI. However, this does not mean getting into the technical details, I am talking about a high-level understanding of the capabilities. The first and perhaps most difficult lesson is that it is not so much that we cannot understand AI, we don’t understand humans very well. As Tom Davenport says: “AI is analytics on steroids”. The creativity and intuition of humans, particularly those at the highest levels of expertise, is still a mystery. The winners of the AI race will be those organisations and individuals who can implement AI with the greatest benefits. For this, executives will need to figure out what AI should be tasked with and what should remain in the human realm. If they get this right, they will achieve the best of both worlds, smart people using smart technology.
KB: How does your latest research approach this? What do the results indicate?
VD: As a practitioner, I was in the AI area for 20+ years. This includes leading AI software development, implementations and also teaching. As a scholar, I have been studying personal knowledge, with emphasis on creativity and intuition. I am particularly interested in how people think and create at the highest level of expertise so I have interviewed top scientists, including 17 Nobel Laureates (see my paper and a blog). Another aspect of this is how people ‘think together’ (see my paper). This duality led me to explore which aspects of human expertise can be rivalled by AI and what is unique about the human mind. I have found that in the areas requiring original thinking, truly exceptional human performance cannot and will not be emulated by AI (see my TEDx talk).
I propose that AI and humans are good at inherently different things. AI is great at processing large amounts of data without making mistakes, which is useful in well-structured, known and stable environments where we have access to complete information and relationships and processes are linear. In other words, AI is good at handling complicated issues. In contrast, humans, particularly at the highest levels of expertise, are performing exceptionally well in ill-structured, volatile and unknown environments where not all the needed information is readily available and relationships and processes are nonlinear. In other words, humans are good at dealing with complexity. This should be the principle based on which we decide what should be assigned to AI and what should be kept by humans.
What did you learn or take away from meeting with the executives at the KnowledgeBrief Innovation Day?
VD: This was another excellent experience in this workshop. Last time, exactly two years before this workshop, I had my first workshop with KnowledgeBrief. That workshop was the starting point of my series of talks delivered globally that culminated in my TEDx talk. This workshop was an evolution from that point and, this time, the audience was particularly interesting for me as there were practitioners who just started thinking about using AI and others that were already using it. This covers a whole spectrum of audiences that I want to talk to – delivering exactly what I have said above in terms of what executives should educate themselves about. It was great to see that they all got my message and that they were immediately able to figure out meaningful use of AI in their respective organisations. I was able to see that my talk made an impact.
With thanks to Dr Viktor Dörfler, Senior Lecturer in Information & Knowledge Management at University of Strathclyde Business School.