By John Glazer
John is an enthusiastic advocate of electric cars and a realist about the likelihood of driverless vehicles appearing on our roads in the near future. He provides an update on where technological progress is taking us. Are we ready to take our hands off the wheel? John is a retired business owner and a member of Dorset Humanists.
Autonomous cars, often referred to as driverless cars, will soon become a reality and not just a pipedream. They are part of the AI revolution which is already changing our lives, and which will lead to a huge number of redundancies. For example, there are about three quarters of a million people who earn their livelihood driving taxis, buses, coaches and lorries in the UK. Within a few years they will all be out of work, at least as far as their present jobs are concerned.
But first, let’s talk about this new technology and how close we actually are to allowing cars to drive us, and not the other way round. Many companies around the world have been working on autonomy for a number of years, and the dream is always “five years away”. However, in the last couple of years, Elon Musk’s company Tesla began training their self-driving systems not by writing traditional code but by feeding video clips into their computers, allowing the systems to train themselves. This is known as “end-to-end” training: video clips in, driving control out. Initially, Tesla faced limitations in computational power, but in recent months they have built an enormous supercomputer believed to be the most powerful in the world, resulting in astounding progress. Progress has been rapid because there are seven million Tesla cars in the world all feeding videos of their journeys back to the Tesla mainframe. This volume of data is orders of magnitude more than that of all the other companies combined, giving Tesla a lead that is unassailable.
Unlike Tesla, which relies solely on cameras for data collection, most other companies have chosen to equip their training vehicles with costly radar, lidar (Light Detection and Ranging), and ultrasonic sensors. In addition, these cars can only work in very restricted high-definition mapped areas. The most successful company using this approach has been Google through their subsidiary Waymo. They now have three, possibly four cities in America where they have a fleet of ‘robotaxis’ taking paying customers around parts of the city. This looks impressive, but is not scalable. It also requires human operators in a control centre able to take control when the car detects an unforeseen object in their way, such as a skip, or roadworks.
Tesla’s ‘Full Self Driving’ cars only use cameras and a neural net to be able to drive anywhere in America. To the naysayers who don’t believe autonomous driving can be done this simply, it is worth reminding ourselves that we drive with just two eyes and a brain, so we know it is achievable. Elon Musk informed us recently that the company had been testing out its robotaxi fleet in the Bay Area of San Francisco, restricted to employees of the company. He also said that there will be robotaxis operating in Texas and possibly California next year, subject to regulatory approval. China has already given Tesla the green light to start testing, and the EU is making changes to legislation to allow autonomous driving, as is the UK. Once the data shows that they are at least as safe as humans on the road, and getting safer all the time, adoption will be rapid.
When this happens, anyone who drives for a living will need to consider their future wage earning prospects very seriously. The commercial pressure on employers to replace human drivers will be enormous. Furthermore, pressure on legislators to approve a technology that saves lives and serious injuries will be immense. All of these vehicles will be electric, contributing to long-term health benefits from reduced pollution in our cities.
While the benefits of driverless vehicles are significant, they come with the downside of job losses. However, history suggests that new opportunities will arise for those who are prepared to adapt and change of direction. We've witnessed similar shifts in our lifetimes affecting coal miners, printers, secretaries, and many other professions. What history teaches us is that such transitions can be devastating for some but advantageous for others. Once a technological shift begins, there is no reversing it, making resistance an exercise in futility.
Eventually, as AI drives mass redundancies across all sectors and Universal Basic Income becomes widely adopted, we could, in theory, enjoy the most valuable commodity of all – TIME. Along the way there will be casualties of the system, but that is the price we will have to pay for progress.
Comments