A child that is five today will turn 18 in 2030. If current predictions prove true, they might not be allowed to control a vehicle.
Given that cars will have significant driver assistance options as standard by 2020, it is fair to imagine that children today will have no desire to control a car anyway. It is easy to make predictions though, so how will we get there?
Solving a problem that people don’t have is not good for business.
While there are certainly many who would prefer to not have to drive, a big reason for many to buy a car is because they enjoy driving.
Trying to convince those people to simply use an automatic gearbox is hard, so convincing them to become a passenger in an expensive product will be even harder.
Instead, the lucrative market that car manufacturers will be focusing on will be the transport market.
Whether long haul trucking, bus transport or taxi services, the demand for these services is high, but the human cost and limitations are a barrier to innovation.
It will not be welcome news for those who drive commercially but the potential revenue for autonomous transport will see these services becoming some of the first to switch. It will first include human monitoring, before becoming fully autonomous - and it could be as soon as 2025 if not sooner.
Will services like Uber be the winner?
Uber's objective has always been to lower taxi costs to the point where you don't need to own a car.
The payment to drivers has been a significant component. Reducing that cost, would allow for more cars to run longer with lower fares. The pressure to switch to autonomous cars will only accelerate if cases brought before courts rule that drivers are employees, not contractors.
Vehicles used in the transport industry will be able to learn a lot about driving conditions and driver behaviour. This will allow for the programming that will be included in personal vehicles making them significantly more effective and safer.
But it will not just be the changes to cars that are required; road spaces and traffic regulations will begin to change too.
Using highways as a starting point, some of the most dangerous behaviour stems from vehicles travelling at differing speeds overtaking slower vehicles. In future, it is expected that trucks - and very likely cars - will be stacked in convoy to both reduce wind resistance and use less roadway.
Should a convoy of cars travelling faster than a convoy of trucks meet, the entire car convoy will either wait for a section of dual lane to pass or execute a passing procedure that is safer than an eager driver believing all they need is an accelerator.
Ban human drivers
For city driving, traffic lights are a necessary evil, but autonomous cars would be able to perform more complex co-ordinated options at intersections. This might result in doing away with them altogether or perhaps switch to making all future options traffic circles.
Which raises the next point. Driverless tech was not just conceived to make driving more convenient, it was to make it safer. And the biggest liability in a vehicle is the driver.
Future laws may be able to get rid of the hundreds of regulations about traffic rules by simply introducing a single new one - ban human drivers.
The simple act is expected to have the greatest impact on reducing traffic accidents and fatalities in the history of motorised transport. Exceptions would be made only for those that have demonstrated an exceptional ability, have advanced training or qualify for specific driving situations.
Young and old drivers are even more likely to crash which would probably mean they would be banned first.
Motor Racing with no drivers
Roboraces are another way to improve the AI systems and might even become a sport in itself.
The first Roborace had just two cars with one crashing before the end, but the fact that it exists suggests this is not just a fad.
While you may think it will still all boil down to the car to determine which car wins; the cars will be identical - just the AI will be different.
And while that may still be some time in the future, the electric supercar called NIO EP9 managed to set a new speed record for an autonomous car - 257 km/h!
For most of us though we happily accept that we don’t want to fly a plane to get to our destination. So, in time, we will probably be happy to do the same with a car.
It does not mean there are no significant issues to still resolve.
The first is how to ensure the navigation system is reliable enough to not fail while in operation. While cars have always been subject to breakdowns, the sophistication and potential difficulty in fixing a fault with a navigation system would make it a significant drawback.
Then there is the complex system that allows the vehicles to learn how to drive, despite them proving to be very stable. This is because operators don’t always know how or why a system learned to do something, and a vehicle could react or fail to react when a human would expect it too.
Besides the autonomous learning, there is also the issue of programmed instructions about what to do if a collision was unavoidable. Who should the vehicle look to protect first? Read more about the moral dilemmas.
The system relies on its ability to connect with other systems online for the most up to date information. But that connection also means there is a possibility for someone to hijack the vehicle digitally.
Recent terror attacks using large vehicles become more chilling when you consider that a successful hack could inflict a widespread and coordinated attack - or that individuals could be held to ransom to release their cars or trucks back to them.
This would be a threat to any component of a digital world but hacked cars are the stuff of nightmares.
Ask a 5-year-old, if they never needed to drive themselves, would they mind? Hopefully, they will say they don’t, because they may never get the chance to anyway.