The OpenOcean CEO Summit, which features speakers from the company’s clients and across its network, was once again a fascinating and in some ways challenging event. It offered not just a deep dive into the future of tech, but also into the way in which tomorrow’s innovations will be funded and developed too.
One of the most prominent debates focused Artificial Intelligence and how it is likely to be harnessed in the coming decade. Open Ocean lined up a stellar line up of panelists including Mike Hyde, Facebook’s Director of Data Science, Peter Lee CEO of Rapidminer and Alex Housley CEO of Seldon.
A wide ranging discussion ensued which touched on everything from the levers behind adopting AI (“generating revenue, cutting costs and avoiding risk”) through to the difference between applied data scientists and coding data scientists and their role in developing AI.
All three panel members hinted that the biggest opportunity in AI would be driven by those who had an understanding of maths, yet at the same time were able to look at what might be coming down the road for their business and how AI might solve those problems.
Finding the right signals
Interestingly Peter Lee spoke about what he perceives as being the number one issue in the state of adoption – finding a meaningful signal. In other words choosing the right data to measure. He gave an example of when an airline recently used AI to see if it could find more effective ways of turning round its fleet. It ended up focusing on external factors like flight time and air traffic control, rather than internal ones, such as preparing the cabin for the next flight, which actually had a more significant impact on airline punctuality.
There was also decision of how AI is only really as useful as its output and the way it could change business. Peter Lee argued that prescriptive action is the key. He exhorted companies to execute data analysis, but also to ensure that the data that was examined was core to resolving business issues.
The panel finished with a discussion about which are the best examples yet of how AI has been transformative for a business. Peters cited a US mortgage lender which in the past had struggled with the key issue of churn. To counter this it has developed Machine Learning that examines a person’s available social data and looks for key factors. The idea being that its customers are more likely to remortgage when their life changes such as having kids, getting married or divorced or separated. Using this demographic data the company was able to approach its customers in real time with personalised emails. This had ultimately generated very significant growth.