Full house at London alumni meeting focusing on algorithms

11 december 2018

The noise level was intense at the Swedish Embassy in London when 60 Uppsala alumni, honorary doctors and guests filled the beautiful, Christmas-bedecked residence. Swedish Ambassador Torbjörn Sohlström and Uppsala University Vice-Chancellor Eva Åkesson welcomed guests to the alumni meeting and lecture on 3 December.

Many had come to recontact with their alma mater, meet old friends and, especially, to listen to mathematician and Uppsala Professor David J.T. Sumpter’s lecture on “The Dangers of Algorithms”.

In her welcoming remarks Vice-Chancellor Eva Åkesson emphasised how pleased she was that so many alumni and friends responded to the invitation to the social gathering at the Swedish Embassy in London.
“To meet today’s global challenges – in medicine, technology, nature, the environment, culture and society — we need good friends. So your commitment and support are most welcome.”

The subject of the evening’s lecture interested and attracted a large audience. Mathematics and algorithms play an important role in many aspects of our lives, David Sumpter noted. Among other things, algorithms can affect what music we listen to, what news we choose to read and how financial investments are made.

Algorithms are used online to direct information to different target groups. An algorithm can be simply described as a process used to solve a given problem. Organisations like Facebook and Google use them to determine user personalities and interests and then specify what advertisements are sent to individuals. If people express racism or sexism in their input, this will be reflected in the algorithms and then show up on our computer screens. This can affect decisions we make as individuals or at the societal level. So it is actually algorithms that learn from people and not the other way around.
“It is often speculated that algorithms and artificial intelligence (AI) are about to surpass people in intelligence, but in fact they have about the same level of intelligence as bacteria,” Sumpter explained.

But more research is required to understand how much and in what way algorithms affects our decisions, Sumpter maintained. Currently a lot of effort is going into developing algorithms, but less energy is devoted to research into their long-term consequences.
How dangerous are algorithms? How far has the research on artificial intelligence really come? Is it possible to create algorithms that do not reflect human prejudices and opinions? Questions from the audience following the lecture in London never seemed to end, and lively discussions continued long into the mingling that evening.