#AI: The hidden algorithms shaping our lives

How artificial intelligence and algorithms influence the information we have access to was discussed in the Parliament on 21 November during the annual lecture of Science and Technology Options Assessment (STOA). STOA provides MEPs with independent, high-quality and scientifically impartial studies and information to help assess the impact of new technologies.

What algorithms are

Algorithms are a set of rules and instructions, usually expressed in computer code, designed to solve a problem. They can be applied to anything from health care to employment.

During his keynote lecture, Professor of Artificial Intelligence at the University of Bristol Nello Cristianini stressed the value of algorithms due to their potential for influencing people's behaviour. “The behaviour of people can actually be predicted simply based on their past behaviour," he said. "We can predict what most of us will do in typical situations. This can of course be used to build machines that emulate people’s behaviour.”

These algorithms affect most of our lives on a daily basis, but most people have little idea of how they work as these algorithms can be very complicated. They can also be biased due to how they were programmed or because of the data used to train them.

The need for transparency

STOA chair Eva Kaili, a Greek member of the S&D group, discussed at the annual lecture how algorithms can influence everyone and everything, for example through the newsfeed provided on platforms such as Google or Facebook.

“If we don’t even know how these decisions arrive from algorithms in our newsfeed how can we judge the quality of these decisions?” she said.

Challenges

However, achieving more transparency is far from the only challenge involving algorithms.

Michail Bletsas, from MIT Media Lab, told MEPs:  "Transparency by itself is not going to solve anything unless you have people who really understand what is going on.” He added: “Artificial intelligence is everywhere, you would be surprised in how many applications it is being used."

Bletsas continued: “We are at the end of a technology revolution and now we have to pick up the pieces and build up on the results of that revolution."

Michiel Kolman, from International Publishers Association and Elsevier, pointed out that machines also had their restrictions: “There are possibilities to fool machines in a ways you can't fool people.”