TU Berlin

TUB-newsportalAn artificial intelligence researcher who does not believe in the concept of intelligence

Page Content

to Navigation

Queen’s Lecture

An artificial intelligence researcher who does not believe in the concept of intelligence

For Zoubin Ghahramani intelligent machines are tools for improving our daily lives. The professor for machine learning and artificial intelligence at the University of Cambridge as well as chief scientist at Uber, will deliver the Queen’s Lecture 2017 at Technische Universität Berlin on 21 November. "TUintern" spoke to him beforehand to find out whether and how machine learning is changing our daily lives.

Livestream from TU Berlin:

An lnterview with Professor Ghahramani

Prof. Ghahramani, machine learning and artificial intelligence are not the stuff of dreams – they are affecting our daily lives right now. The term artificial intelligence raises many concerns in society. Where exactly does the difference between machine learning and artificial intelligence lie? 

From an historical perspective, machine learning is rather a spin-off of the classical science of artificial intelligence. Machine learning consists of equipping a computer with an algorithm which enables it to recognize and learn patterns from colossal quantities of unsorted data. In practice the borders between the statistical analysis of data, machine learning and artificial intelligence are very nebulous. We can also see this in the fact that machine learning appears in many practical areas which we do not perceive of as “intelligent” at all. The process of a bank checking to see if the use of a credit card abroad is an attempt to commit fraud or not is based on statistics and machine learning processes, but we would not describe this as artificial intelligence.

Zoubin Ghahramani, professor for machine learning and artificial intelligence at the University of Cambridge and chief scientist at Uber

What exactly are then intelligent machines?

Although I originally come from the area of artificial intelligence research, I do not believe in the term ‘intelligence’. The problem is that we, as humans, tend to think in a very anthropocentric way: Everything we can do well, we call intelligence, while the things we do less well are not seen as intelligence. Sure, I admire my calculator for being able, unlike me, to multiply very large numbers in a matter of seconds, but is this intelligence or a ‘super-human’ ability?
We have to break away from our human-centered notion of intelligence. We are a type of primate able to do certain things very well. Many animals are good at other things and computers are able to solve certain problems effectively. The interesting question from a scientific perspective is not: ‘What is an intelligent machine?’ but rather: ‘Who can solve a particular problem and to what standard?’. We can then compare and find out who best solves this problem: a human, an animal or a computer. Seen like this, then human intelligence is just the name we give to a set of skills. Computers or animals possess other skills. Intelligence is merely a word and not something substantive.

What do you say to people who are scared that intelligent computers will shortly assume world domination, making human work or even humans themselves redundant?

Machines – including intelligent machines – are tools intended to make our lives easier. The only real reason to build them is to ease the burden on humans. A modern smartphone provides us with “super-human” skills. The more such machines we have, the more we should see them as something which makes our lives easier and not something to replace us. Machines are not vain, envious or power-hungry. Of course, intelligent systems can be used to manipulate people. But just as easily, we can use them to uncover such manipulations.

But: Increasing automation and the use of machine learning processes will dramatically alter our working world in the future. We need to consider this. Which tasks, which jobs will be done by computers in the future and how will we deal with this? I sit on a number of interdisciplinary committees which research the influence of artificial intelligence and machine learning on our daily lives. The goal of this research has to be to ensure that the impact of the advantages and disadvantages that the use of such systems brings with it are equally and fairly distributed throughout society. In general, though, I tend towards a positive view of things: Looking back over the past few centuries, we can see that, thus far, the use of technology has improved rather than worsened our lives.

Many thanks!

Interview by Katharina Jung


Quick Access

Schnellnavigation zur Seite über Nummerneingabe