Few businesses are familiar yet with what terms like data scientist or machine learning actually mean, says Gabriel Straub, Head of Data Science and Architecture at the BBC, and it is even less clear to most people how new technologies will affect our work and world.
This means that hiring the right people is tough. What does a data scientist or expert in machine learning actually do? Straub says firms like his former employer, Tesco, have been collecting lots of data and finding trends to predict future behaviour for years. Now, the scope and scale of th e data and ability of machines to process it and make predictions is a massive opportunity.
Data scientists, he says, need to know the problem they are solving. For example, at the BBC they ask: “How does data science help me to watch a better TV show? How do we create a more engaging experience – a BBC that wraps around your life – rather than you having to change your life in order to consume the content that we have?”
Nicos Savva, Associate Professor of Management Science and Operations at LBS, says business leaders don’t need to understand all the data but they do need understand the business problem that the data scientists can help to solve. “Become the link that translates the problem and implements the solution.,”
He also examines the concerns arising from removing human judgement from an automated process. “There are two challenges. Bad people will do bad things and it enables them to do them at a larger scale and more sophisticated manner. A classic example is a company targeting ex-gamblers with gambling ads. The other challenge is that we don’t understand the limitations of data science and we tend to rely on recommendations too much. An example is credit scoring. It is designed to work well on average but it’s quite possible that somebody gets an unfairly low or high score.”
Human intuition, says Julian Birkinshaw, Professor of Strategy and Entrepreneurship and Academic Director of the Institute of Innovation and Entrepreneurship at LBS, remains as important in business as it always has been when it comes to making the best use of what machine learning can do for organisations.
Previous podcast: Narayan Naik, James Dunne and Julian Birkinshaw ask whether rob-advice can replace professional fund managers.