I have been attending the AI & Big Data Expo Europe Virtual conference. Here are three speakers who stood out for me.
Implementing Robots at scale, a case study at ING – Jeroen van Genuchten
I especially liked his explanation of how they use the same identity management sytems for employees and robots to manage access rights. It shows how you can integrate robots into your existing work processes.
Becoming a Data Driven Company, Data at the Core of the Business Ecosystem – Gillian Tomlinson
Her definition of data-driven is having data points across the organisations that drive decisions. I especially liked the value driver framework that she presented, that defines metrics to measure the company value levers (e.g. increase revenues) and drivers (e.g. sales to existing customers).
Then she maps business initiatives across the value drivers to check if the initiatives fit their strategy – increasing revenues vs reducing costs for example. Good stuff.
Fairness & Explainability in AI – Javier Campos
Javier Campos gave an excellent and crystal clear explanation of how to get fairness in AI. All business leaders using AI in their companies should learn about it.
He explains that fairness is a complex concept that depends on the context and differs per stakeholder. What is fair for a consumer may not be fair for a company.
He then presents the “fairness impossibility theorem” – there are many different ways to measure fairness and you cannot meet all of them at the same time.
And finally, not using a particular parameter in your model (e.g. gender) doesn’t prevent it from being unfair.
© 2020 Veridia Consulting