It is hard not to notice the surge in interest in producing driverless technology of late. It seems most car manufacturers and plenty of big technology firms are desperately trying to get their PC’s on wheels into the public domain as quickly as possible and whilst some countries and governments have rallied against idea (India a prime example) they are definitely in the minority. A far greater proportion of countries and governments however seem to exhibit considerable excitement at the prospect of autonomous travel.
I wonder to myself why this may be, when I think it is fair to say there is pretty low demand for driverless technology among the general population. I would almost go as far as to suggest that some of us actually enjoy driving so why would we want computers to have all the fun? Where is the driving force for this technology really coming from (excuse the pun)? Well, let’s consider the amount of money logistics and delivery firms could potentially save not paying human drivers wages. That would surely be huge, the incentive there alone must surely be enough to convince any capitalist profit hungry mega corporations with a glint in their eye for a few more billion in the bank (or the off shore accounts). Let’s be realistic, pretty much all the demand for this technology is being generated by the entities that stand to make the most money from it and they will throw vast sums of money at capitalist and any other governments open to donations to support their cause – thereby providing the framework for the prophecy to fulfil itself.
Over the course of my life I have had cause to use computers a great deal, having experimented with programming in a variety of languages and seeing the technology rise from being ‘something for geeks’ to becoming something that is intrinsically woven into the fabric of our lives. Computers are without a doubt really useful, in fact in many ways they are essential and we would struggle to do lots of things without them, but would I trust a computer and it’s man made algorithms in an autonomous driving vehicle on the road with other human drivers? Not a chance.
I can envisage countless scenarios where humans will ALWAYS be the most capable of making the best decisions in a split second when compared to algorithms and a set of electronic sensors. It’s not just the obvious stuff like what is around you, driving is so much more than that and the argument that it will make the roads safer is utter nonsense. Unless you remove all humans from the roads (and I can guarantee that it something that will never happen) there will always be an element of unpredictability and algorithms can only cope with so much unpredictability. We encounter it so much on a daily basis that it’s become very much instinctive as to how we deal with it.
Consider if you will, the other senses that we have at our disposal, the ability to hear things coming – I’m sure driverless cars can and may already be fitted with sound sensors, but the amount of data they may need to process at any one time, it is ridiculous to expect a computer to pick out a hazard from the noise around it. How will an autonomous vehicle know a beep from another car near it is not intended for itself? How will it know that the car that has just flashed its lights at it is saying a ‘thank you’ rather than there is a problem with your headlights? I mean we struggle enough with these sorts of scenarios let alone leaving it to Windows 95 on wheels!
Of course it’s not just our sight and our hearing, we also have highly refined sensors to ‘feel’ and react to all sorts of vibrations, from the smaller ones; such as when you can feel a strong gust of wind blow your vehicle into the path of a large lorry a long way ahead of you. With a human driver of said lorry you will find that most capable lorry drivers are quickly able to apply the appropriate reaction to counter this effect. But you need to be ready to react as well, and as a human you are already prepared for the possibility. A computer driven vehicle is making decisions based on data that is to all intents and purposes incomplete and at times irrelevant. The small vibrations are one thing but when an autonomous car does have an accident and crash into something (note there are no if’s here!) how can its mixture of damaged and still functional sensors detect the most appropriate response to ensure the best possible end result in terms of safety? I believe the proponents of driverless technology hugely underestimate the adaptability of humans in comparison to computers. It is no secret that most accidents are caused by humans, that’s never going to go away but putting soulless monolithic technology with vague accountability into the mix of real world drivers definitely won’t improve that, if anything it will make things worse!
When I say driving is so much more than vision that is a massive understatement. There are so many subtle nuances that come with driving in the real world that just can’t be picked up by computers, the barely detectable hand gestures or facial expressions from other (human) drivers, the moments in driving situations where nobody is sure who has right of way to go first and either everybody goes at the same time causing a gridlock, or it is up to someone to go first and hope for the best. The roadworks, the pedestrians, the drunk pedestrians, the temporary traffic lights, the animals running across the road, the damaged parts of the roads, the natural disasters and sinkholes, you name it, we know how to deal with it better. The roads in the US are one thing, long and wide a lot of the time, lots of space and more than a few half decent places to try out this technology, but if you have ever driven in Europe there is level upon levels of complexity to add to this and to take into account. Smaller windy roads, tiny carriageways that only have space for one vehicle at a time to pass through in either direction at various points along their length, and we haven’t even mentioned the ‘r’ word!
If you have ever driven in Europe and navigated a simple roundabout then that’s one thing but in the UK there are these unique things such as double roundabouts, even triple and quadruple roundabouts! There are places in Europe where the roundabouts are so fiendishly complex to navigate your way around that even with advanced electronic sensors to match most human senses I can envisage most driverless technology experiencing the driving equivalent of the Windows Blue Screen of Death! If I could pick one example that is like no other, go and see the Arc de Triomphe in Paris at any busy time of day (which to be fair is pretty much all the time anyway) and prove to me that any computer in the world could cope with that, let alone do so consistently! It is pure and utter chaos on a grand scale, everyone taking their lives in their hands at any given moment and we actually want to entertain the idea of adding stupid driverless vehicles into this madness?! It’s asking for trouble but we give computers too much credit to expect them to cope. They can’t and they won’t.
Here’s my humble take on how I expect driverless technology to play out for the next 5-10 years: the proponents of it (have) and will come to realise that it is much more complicated a problem than they first thought. But by the time Uber, Amazon, Google and all the other firms peddling these wares realise this it will be too late and in some places around the world very real and terrible accidents are going to occur and there will be plenty of legal wrangling over accountability. Some of these firms are going to find themselves facing some significant legal bills and whilst there will be some places where the technology can and will work reasonably well, there will be horror stories coming in from other places and reputations and share prices will be made or broken. And let me also throw in the huge protests that will come with this technology. To the enterprising capitalist keen to pay the least possible price for deliveries, logistics or human transportation, once this technology does start to gain traction there will be huge job losses in industries that employ millions around the world. These people, who rely on their ability to drive to pay the bills are not just going to disappear quietly and go off and find new careers easily, you can expect a very hostile reaction from them and understandably so.
We’re still in the early stages of letting this monster out of the box, there are various ways this could play out and right now we have still have the ability to shape our own destiny with this to some extent. But as a society and as a species we have a duty to ensure that we do not just arbitrarily accept the notion that computer controlled vehicles will make us safer, and we should certainly question the motives and the need for this technology in the first place. In our race to create the most advanced machines to do more and more demanding tasks we forget that the most advanced being in the (known) universe that requires the least power and fuel to do the widest range of tasks is reading this very sentence.
The site is called collective responsibility and I think in this area, like many others we do have a responsibility to do what is most sensible and responsible rather than what can generate the most profit. My opinion, for what it’s worth is that rather than trying to integrate the world of human and driverless vehicles, why not create new roads for the driverless vehicles and let them only have to deal with other computers that (one would hope) have far greater predictability than humans and could also offer a genuine possibility to not only be safer and just maybe even save on journey times! I have not come across many other mentions of this potential solution in my reading elsewhere on this topic but I would certainly be interested to know if others are making similar propositions? Yeah there is no doubt this approach is going to cost more, but might it be a price worth paying for getting this right?