Merc R&D chief Ola Källenius on why artificial intelligence doesn’t mean the robots taking over
> IN MY job, you tend to be a technology optimist by definition. On the cusp of this knowledge society, supported by AI, job profiles will change and companies and individuals alike will have to adapt.
> IF WE keep that mindset, I believe there will be an explosion of new areas to be explored within the realms of technological possibility.
> WE’RE SUDDENLY talking about AI in such an aggressive way. We were watching movies back in the ’90s like Stephen King’s Lawnmower Man thinking we’re going into a robotised world, but then for 20 years it went quiet.
> WHY HAS it come back now? Because a couple of things have happened and one is absolutely crucial: computing power. We’re using AI to develop autonomous cars – if we wanted to do that five or 10 years ago, we would have had to hire a truck to store all of the computers needed to make these calculations. With the new processors that are available to us, you can package it in a small box. There’s a huge increase in computing power and sensing technology.
> WITH THE artificial intelligence we’re using now, it’s like the engineers woke up in their sandbox and had new toys to play with. We’re on the cusp of getting the autonomous car problem solved.
> WE’RE REALLY just at the beginning of understanding the potential of artificial intelligence. For example, my smartphone tries to get to know me through my behaviour and make suggestions. We’re trying to put this into our new connected cars; they’ll learn user patterns using simple AI. These are just baby steps and there is a lot more to come.
> IT DOESN’T replace original thought, though. I’m not one of these Skynet pessimist who think that computers will take over the world, but I can see an enormous amount of potential in it as an intelligent assistant for new technologies – and these technologies will be for the betterment of all of our products and the company as a whole. For example, one of the main reasons why people buy a car is the way it looks – this is still very much the case. But deciding the aesthetic beauty of a technical object is down to the human mind only.
> WHEN WE design our new cars, the buck stops with a few individuals who say that they believe this is the design trend five years from now. That’s all gut, emotion and experience as well; I don’t think there’s an artificial intelligence algorithm that can solve that yet. If you’re wrong it can cost you a lot; if you’re right you can have a tremendous amount of success. INTERVIEW BY JAKE GROVES
FRESH THINKING JLR’s seeing autonomous cars
If looks could kill… no, wait, that’s not the point
Another autonomous pod?
Look a little closer. Along with the cameras and radar, there’s something that looks a little diferent.
Wait, are those eyes?!
That is correct. JLR has been working on this as part of the huge self-drive test programme it currently has in progress. That’s mostly about flooding Milton Keynes with Range Rover Sports with Level 3 and Level 4 autonomous capability (the car can drive itself but the driver needs to be ready to take over), but also includes more radical ideas like these ‘virtual eyes’, which look at pedestrians the way a driver would.
Why do I want a self-driving car staring at me?
JLR points to research that says 63 per cent of pedestrians are worried about safety when crossing the roads of the future, so it’s worked with cognitive psychologists on the concept of trust.
So where does 2018’s Johnny 5 come in?
It’s part of an experiment to measure trust levels before and after the car ‘looks’ at you. So in practice, when it notices you waiting at a zebra crossing, for example, you’ll be confident about crossing because the eye contact has assured you that you’ve been seen.