There are too few data scientists in the world and education needs to change in order to maximise the true potential of data science, one of the leading authorities in the field has told Computing.
"The answer is 'no', there aren't enough data scientists, not even close," Alex ‘Sandy' Pentland, director of the Human Dynamics group and the Media Lab Entrepreneurship Program at the Massachusetts Institute of Technology (MIT) said, when asked about the pool of data scientist in the world.
Earlier in the day, Pentland had given a keynote at the O2 Arena, as part of Campus Party Europe, in which he told the audience the data revolution will dwarf the internet revolution. But in an interview with Computing, he said in order to maximise that potential, the way data scientists are trained needs to change.
Pentland argued that more data scientists need to come from an electrical engineering background, rather than mathematics.
"One of the things that happened, which is sort of unfortunate, is most of the computer science type of data people are trained in a type of mathematics called discrete mathematics... [which is not best suited to] the sort of data that we're collecting now," he said.
"The people who are much closer are the electrical engineer types, who are used to signals and audio and video and stuff like that. But they also have to then extend themselves to become much more familiar with machinery and human behaviour," Pentland continued.
"And so the most effective data scientists have a background in things like signal processing, and then we're into machine vision and finally... human behaviour."
Pentland argued that in educating data scientists, society itself needs to be treated like the quantitative science he believes it has become thanks to the proliferation of data in the world.
"There needs to be general literacy about data interpretation and also much more literacy about the way society works," he told Computing.
"Because the way society works is now becoming a quantitative science, it's not the part of humanity that theorises about psychology and so forth. It's actually now data science and people need to become more familiar with some of the lessons about that.
"For instance, the importance of social learning, learning from each other, that creates fads, that creates market crashes," Pentland continued. "We tend to teach people that everything that matters happens between your ears when in fact it actually mostly happens between people."
Pentland added that data and society will go hand-in-hand and that that message needs to be taught from an early age, as it'll ultimately aid data science.
By eliminating high entry costs for big data analysis, you can convert more raw data into valuable business insight.
A discussion of the "risk perception gap", its implications and how it can be closed