Although technical progress may seem invincible and here to stay, control is not out of our hands, says Professor Bernd Neumann, CEO of the Hamburger Informatik Technologie-Center e.V. (HITeC) at the University of Hamburg. In an interview with Hamburg News, Neumann gives more insight into artificial intelligence.
Hamburg News: Professor Neumann, when is a computer programme intelligent?
We speak of artificial intelligence when a programme does things that require human intelligence. The exact moment is unclear as we are dealing with a smooth transition. We still resort to a modern version of the Turing test to determine that moment. (The Turing test was developed by Alan Turing in 1950, as a test of a machine’s ability to exhibit intelligent behaviour equivalent to or indistinguishable from that of a human.) Artificial intelligence has yet to master the challenge of humour the understanding of which requires not only pure knowledge, but also that of a cultural background. In actual fact, artificial intelligence cannot be viewed as a miraculous elixir, but rather as the consistent application of automated technologies.
Where can such consistent applications be found in future?
There are three main fields of application: Turning big data into smart data by extracting valuable information from a large amount of data. Then there are autonomous systems i.e. machines take over complex tasks and lastly, virtuality or simulations of complex technical tasks on a computer. We focus on concrete projects in such fields at HITeC – the research and technology transfer centre in the Faculty of Informatics at the University of Hamburg.
You are CEO of HITeC. Can you give examples of some projects?
We implemented the Hamburg law on transparency, effective from 2012, in co-operation with the Hamburg Revenue Office and prepared a project using artificial intelligence to evaluate traffic data. We also focus on monitoring traffic scenes. The artificial intelligence application first learns to distinguish people, cars and bicycles from each other. The next step is interpretation to identify possible traffic hazards. We undertook a project like that in co-operation with Toulouse Airport. Our task was to monitor and interpret events at the apron. We work on a whole series of different projects in co-operation with science and commerce at HITeC some of which are funded by the EU, the Federal Ministry of Education and Research, the Federal Ministry for Economic Affairs and Energy or the Innovationstiftung Hamburg.
Where do we stand today – will artificial intelligence surpass human intelligence soon?
That has already occurred in certain sectors. Just think of IBM’s Watson and the advances in autonomous driving. And in many other sectors, this development is unavoidable which is causing many people headaches. The British physicist and astrophysicist Stephen Hawking views artificial intelligence as a threat to humanity.
And what do you think?
I personally think that artificial intelligence can improve our well-being and may go a long way towards ending wars as well as alleviating poverty and pollution. And I am convinced that artificially intelligent creatures can be controlled, if we wish. Bright minds today and in former times have had many heated debates about this problem and the associated fears. In the early 1940s, the sci-fi author, biochemist and visionary Isaac Asimov formulated laws on robotics the most important of which was “a robot must preserve mankind”. And in view of this complex technology, that principle still holds true today. The control of artificial intelligence must take account of human values.
The value of people in work processes is under threat. Is artificial intelligence forcing us out of the labour market?
That is to be expected. However, I think that is good in some sectors. Artificial intelligence can deliver results far faster and more precisely than humans in many sectors e.g. the diagnosis of rare diseases in medicine. Artificial intelligence is far more effective and our economy adjusts to promising technologies as soon as their effectiveness has been proven. The challenge now lies in planning the long-term control and development of artificial intelligence and avoiding unemployment. That has to begin with training and the standards for professions must be adjusted.
A last glance towards the future – what’s next in artificial intelligence research?
Developing emotional robots is the next, most interesting step. Artificial intelligence must learn to understand and simulate human emotions. That requires a combination of understanding language and facial expressions. A car driver could give a digital assistant a voice command such as “radio on” and, artificial intelligence would search for a radio station based on facial expressions until the right one has been found. However, this application would be particularly interesting in nursing where it is a hot topic at present.
Interview by Yvonne Scheller