Raúl Rojas from FU Berlin

Raúl Rojas is Professor for Computer Sciences at Freie Universität Berlin and has carried out researches on Artificial Intelligence (Künstliche Intelligenz [KI]) for forty years. The Mexican-born professor became internationally known by his robots playing football - the FU Fighters. Currently, Rojas is working on self-driving cars, whose prototypes are called ‘Spirit of Berlin’, some of which are already driving around Berlin.

In your opinion, robots should take over the steering wheel in the near future as they are better drivers compared to human beings - and this is to make traffic safer? In future, computers will be able to drive safer than human beings. These computers will monitor the traffic all around with laser scanners and video cameras. In addition, these cars will communicate by radio with each other. A computer does not get tired and can brake extremely fast. How can real people keep up with this? It would be quite easy to control current traffic with robots but the factor human being is the problem here. Until computers have learnt to behave like human beings at the steering wheel and how pedestrians react, it is not possible to use these robot cars in cities. Robot cars will first of all conquer motorways and only be permitted to drive in cities after many years of testing.

Seen on the long run, traffic will become safer but above all, we will have fewer cars. The car will change into a taxi, and in combination with public transport, we will be able to move the whole city.

A robot will subsequently be more intelligent than the human being that created it…? Not at all. Artificial intelligence means that computers will be equipped with certain abilities, as for example, playing chess. In this very restricted section, computers will outclass human beings - however not in respect of really understanding matters and complex discussions. Computers are ‘idiot savants’. The same as a locomotive can carry much more than a person or a steam-driven hammer can reshape metal with more force, a computer can surpass human beings in special areas without actually knowing it. The computer remains an instrument, our extension, a ‘dumb’ prosthesis.

When will your vision become reality? The first prototypes are already driving around some cities. Starting in 2020, autonomous vehicles will be sold for motorway use. However, the driver must always be ready to interfere at all times. The practical use in this case is still rather low. Progress is developing very slowly at this point: The technologies are already available resp. will be available within a short period of time; however, the legal and social basis is still missing.  Subsequently, we are talking about decades, as the cities will have to be planned for this situation resp. be prepared for the change.

Tesla boss, Elon Musk, has reached his limits with his so loudly announced robot cars… In the past there had been some accidents - rather serious in part - , not only at Tesla but also with the mobility service provider Uber – are you still looking ahead with confidence? Tesla as well as Uber went into the field rather aggressively, without installing the required safety measures. It is very important to progress cautiously, carry out tests, and more tests such that nobody will be in any danger. There have been no accidents, for example, in Germany. Companies here are much more prudent as they have to protect their reputation. If a company such as Tesla and Uber have experienced losses year after year, one does not care too much about the reputation but more about showing off.

What will manufacturers have to do to avoid such accidents? Have a schedule, maintain the necessary safety standards in order to programme cars as a whole, like a completed work of art. It is, for example, farcical considering that the accidents at Tesla and Uber could have been avoided by simply combining the information coming from various sensors.

How can or even must politics use artificial intelligence? And where should robots not be used? I would not want robots as a carer or so-called ‘care givers‘, or to be used to replace soldiers. Once armed conflicts are delegated to machines, the readiness to fight again will become easier or existing conflicts may easily escalate. Computers cannot accept responsibility - only human beings can do that.

Politicians should not use KI. They should think more about how KI will influence our privacy, what consequences the employment market will suffer, and in what way social isolation would develop. I am always rather critical towards social networks as they misinform rather than inform in most cases.   

How many robots do you have in your household? Only a vacuum cleaning robot - rather an oddity working its rounds. I am much faster when vacuum cleaning myself, perhaps even 20 times faster than the robot. And I burn off some calories in this way. ​​​​​​​

You have been living in Berlin since the 1980’s - what made you come here? Doctoral studies first of all, followed by one project after the other. Before I thought about it, ten or twelve years had passed and I stayed and built up a scientific career in Germany. I chose Berlin for my doctorate as it appeared to be a fascinating city looking at it from far away. And it proved to be true.  ​​​​​​​

And finally: Would you please complete the following sentence: “Berlin is smart, because... it is multi-cultural.