The reality of the driverless car
Back in 2014, it seemed inevitable that autonomous cars would be the vehicle of choice within a matter of a few years.
A lot of people predicted that we would be “driving” them by 2018, then they would be dominating the roads by 2020. Now, even the most optimistic are saying they won’t be used by the general public until 2028.
Some are saying it may never happen.
The most famously successful autonomous car was Chitty Chitty Bang Bang, the second was arguably K.I.T.T. - driven by David Hasselhoff in Knight Rider. You could be forgiven for wondering if cars having the ability to be truly autonomous is just as much a work of fiction as these stories themselves.
Why might self-driving cars be a non-attainable achievement?
Elon Musk explains that our real-world environment for autonomous cars is designed for people to make unsupervised full self-driving, using biological, neural and experience skills, to perform the driving functions as well as the awareness functions. But it hasn’t been possible to program computers and electric cars to do that.
Many believed that the full-self-driving feature on Tesla’s would be working by now, but it has proven impossible to implement.
The invention of a gaming theory called GAN, (general adversary network) simulates different scenarios to play out that would allow a machine to learn what was going on around itself.
Look at the scenario work out the solution, put that solution in motion. However, GAN in autonomous cars simply hasn’t been able to accomplish the complexity of driving, particularly in cities and built-up areas.
Why is this?
As humans we go through a cycle of learning. Also known as the “Cycle of Consciousness” – to read more about this check out this blog post.
Humans are able to reach a point where not only can we drive a car, but we can hold a conversation, listen to music and (although strongly discouraged) smoke a cigarette! Which just goes to show just how advance humans are in comparison to any form of autonomous vehicle.
Therefore, at this moment in time, it seems unlikely that the current technology will be able to get autonomous cars to be working in places like cities where the variables are just too expansive.
Self-driving cars could work on motorways really easily. As long as you have a common agreed way of the vehicles knowing where they are, because everyone is travelling in the same direction, there is less to think about.
You could even have the technology in place for the roads to talk to the cars! In theory, driverless cars are almost certainly suitable for motorways.
The minute you enter an environment where there are so many variables, from every possible angle and the behaviour of all of those variables in not entirely predictable it becomes a hell of task for the vehicle to work out what happens next.
Even if you were able to get a car to the point where it could see all of those possible hazards, it still puts the vehicle in the position where it has to make choices, which it is ill-equipped to do.
According to Asimov’s Three Laws of Robotics, a robot:
- May not injure a human being or, through inaction, allow a human being to come to harm.
- A robot must obey the orders given it by humans beings except where such orders would conflict the first law.
- A robot must protect its own existence as long as such protection does not conflict with the first or second law.
Whilst we may need more nuanced interpretations the issue with the autonomous car is that if it has no way to avoid hitting something and if it had a choice between hitting a child or woman walking a dog, how does it make that moral choice?
Also, who decides which choice the car made was the correct choice, and who is responsible for the consequences of the cars actions if someone was to come to harm? A human would have their right to make their own moral choice and may even choose to sacrifice themselves in order to try and save the child and the woman with the dog. But the car can’t make that choice because if it sacrifices itself, it’s also taking a human (the “driver”) with it.
So one of two things needs to change - we either invent completely new technology because our current technology is too limited or we invest heavily in infrastructure in order to make the city a safe place for an autonomous vehicle.
It’s not just the technology
In addition to the advancement needed in technology and infrastructure, the legislation requirements in order to manage and hold people accountable for the emergence of this type of technology would have to be constantly evolving to be able keep up with robotics.
It was only in the past couple of years that London’s airports came to a standstill due to drone activity in the area, quickly resulting in new laws about where people could fly their drones.
The social morays would also need to be considered. It is currently illegal to drink and drive, would this be an issue with driverless cars?
There are countless questions that need to be answered before autonomous cars become anything like a dominant form of traffic on our roads!
In the meantime, you take the wheel!
Did you enjoy this blog post? Subscribe to our blog to be updated when we upload new content!
