Bram Schot on the impact of AI on mobility
Marieke Martens, science director at TNO and professor of automated vehicles at the Eindhoven University of Technology, talks to Bram Schot. Schot was the CEO of Audi until 2020, having previously held management positions at various car makers, including Mercedes and Volkswagen. Their conversation concerns the influence of AI on mobility. How will AI impact the production process? And what does a future with autonomous vehicles look like?
A conversation with Bram Schot
What do you think of AI?
A lot of people are talking about the huge potential of AI, but others have in fact become more cynical about it. So where on that spectrum are you located?
‘On both sides. In my opinion, the possibilities of AI are unlimited. But if you don't tackle it in the right way, there can also be a downside. It’s a matter of finding a balance. So that leads to some questions: Who will regulate AI? What are we going to do with it? What are the goals? It’s very important that we talk to one another about all that. If humanity benefits from it, then the question remains “What is humanity?” Because you also hear it said that AI encourages inequality. In the case of the Dutch childcare benefits scandal, someone decided to put down a few parameters on paper, based on which the system was going to make certain choices. We can of course look at the system, but that was really a case of human failure. I still see AI primarily as something that’s supportive of humans rather than taking over the role of humans.
‘We humans are also biased. Nobody is objective. When I prepare to talk about AI, I can find just as many articles saying that it’s great as articles saying it isn’t. If AI could help us get back to what really are the facts in this world, then that would be very welcome. I believe the world must be on its guard about slowly but surely losing track of the human dimension, about looking only at the figures and no longer understanding the world behind them.
'If we no longer see the world behind the figures and also don’t have any time left because we’re busy analysing and collecting data, then the world will slowly become artificial, both literally and figuratively. I believe there’s an incredible need to regulate certain things better in this world. And I think there’s also an incredible need to slow down a bit and concentrate more on quality than on speed.’
'You have to strike a balance somewhere. What if I give up a bit of my privacy, but on the other hand I get something in return?'
An issue that a lot of people are concerned about is the self-driving car, autonomous driving. How do you see that?
'It's going to happen at some point, but not as soon as everyone thinks. Car companies are currently working on digitalisation. They’re working on electric vehicles. They’re working on hybridisation. They’re working on autonomous driving. They’re working on new business models. They simply don’t have the budget or capacity to do everything at once. And they often don't have the skills to do everything at once either.'
'Perhaps we should use the introduction of artificial intelligence to put things on hold for a bit, and to press the reset button.'
Is that a change compared to a few years ago, when there was less pressure in the area of sustainability?
‘Yes, I think it is. There's no question that electrification needs to happen. It has to happen because the available technology for meeting the emission values is mandated by the legislature. There's no way around that. A number of companies have already had to make two or three restarts. And you may well wonder whether it makes sense for three different car companies in the same sector to try to reinvent the wheel independently of one another.'
Should we be in a rush?
A few years ago, the battle was about who would be first. That’s perhaps because everybody was in such a rush at the time. Is that really how we should tackle it?
'What does the customer actually want? Initial studies indicated a “take rate” of 15% to 20%, depending on which generation you asked – whether that was Generation Z or the millennials. So you already get different answers. Ultimately, the important question is what you solve with it. We currently have 1.4 to 1.5 billion cars in the world that are just standing there unused 85% of the time. And when a car is moving, there are on average 1.2 people in it. That's not what you'd call efficient use of the car fleet.'
'Why deploy autonomous driving for 100% of the possible uses? First start practising where we can set conditions for it.'
So the solution is basically: more people in a car. To get back to autonomous driving: what will it solve?
'What’s your definition of autonomous driving? Based on business models, it’s sometimes said that if we can get rid of the car driver or taxi driver, we can greatly reduce costs. If that’s your only approach, then yes, you will in fact reduce costs. You’ll perhaps make fewer mistakes. But it would be a pity to use autonomous driving only to achieve that. We simply have too many traffic movements. And if we went from having 1.4 billion cars to having 1.4 billion autonomous cars, we would have solved nothing. If we can simultaneously ensure that we go from 1.2 people per car to four, five, six, seven, or eight....'
'If you try to develop autonomous driving for all the possible uses out there in the world, it becomes pretty difficult. But on a remote site, in a newly designed city, with separated lanes, then it’s quite possible at lower speeds. I think we should embrace it, especially if we also make sure we go from one to four people per car, or to a robotaxi with eight or ten passengers, also, in particular, to reduce costs. Because eventually, four or five years from now, mobility will be unaffordable for a very large part of humanity, definitely if that mobility means cars. There's a new generation for whom it’s a matter of principle to spend less on mobility anyway. So a self-driving vehicle that carried more than the average 1.2 people would be a great solution.
'It already starts with the question of “how fast do you let a vehicle go?” If you say that you're going to deploy autonomous driving wherever it actually delivers the greatest benefit and involves the least risk, then we will already have achieved a great deal. But why deploy autonomous driving for 100% of the possible uses? First start practising where we can set conditions for it, and where the risk is really extremely small.'
At TNO, we are discussing the ethics of AI. What’s your view on that?
'I've been to China a lot, and there you see two sides of the coin. You see all those young people there, with their ambition, drive, and energy – that really drives innovation. On the other hand, we Europeans think differently about privacy, for example. I remember being on a visit there and one of them asked “Bram, can we take your photo? Then we can tell you where you've been the past seven days.” Then you can simply see – with amazing accuracy – exactly where you've been over that period. So there I'm giving up a great deal of my privacy. But at the same time, if you look at the crime rates in China, in this case in Shenzhen, they’re extremely low. Those are considerations that you have to weigh up.
'You have to strike a balance somewhere. What if I give up a bit of my privacy, but on the other hand I get something in return? I think that's the way we need to go. I’m not at all bothered that people know a lot about me, or that they have a lot of data about me and use it to make a profile of me. I don’t have a problem with that, but there’s a limit to what I think is OK. There is a way to specify a minimum limit, and that’s a discussion we need to have with one another. Right now, I don't think that discussion is taking place. And that's because we live so fast, in a big black box. We take so many things for granted that we can’t find the time and the peace and quiet to have those discussions with one another. Perhaps we should use the introduction of artificial intelligence to put things on hold for a bit, and to press the reset button.'
Download vision paper
Download vision paper ‘Towards Digital Life: A vision of AI in 2032’
More about 'Towards Digital Life: A vision of AI in 2032'
- David Deutsch on the development and application of AI
- Georgette Fijneman on the promise of AI for health insurers
- Rob de Wijk on the rise of AI in geopolitical context
- Eppo Bruins on AI in different government domains
- Bas Haring on AI, science and philosophy
- Arnon Grunberg on AI, creativity and morality
AI Systems Engineering & Lifecycle Management
The AI system for the future. At TNO, we work on AI systems that remain reliable and can handle new functions in the future.
Responsible decision-making between people and machines
Bias in facial recognition and accidents with self-driving cars. AI must be developed further. The fastest way to do this is in close cooperation with people.
Knowledge representation and reasoning
Correct and unambiguous information is needed when making a decision. That is why we use AI technology called "knowledge representation & reasoning".
Natural language processing
What is natural language processing (NLP) and how do we use it intelligently? Find out how we use this AI technique to gather information from textual data.
Robotics and autonomous agents
Robotics brings a future-proof industry a big step closer. For example, we are working on automatic path planning with AI techniques.