Discussion Of Disadvantages Of Self Driving Cars
Self-driving cars are developing rapidly due to automation and artificial intelligence and are becoming more common on the roads. According to a survey by the US Department of Transportation's National Highway Traffic Safety Administration, 94% of vehicle crashes are caused by human error. Self-driving cars are claimed to be safer as they eliminate human error and can reduce distracted and dangerous driving-based incidents. Nevertheless, some accidents cannot be completely avoided and when accidents happen, who should be responsible? For instance, when a self-driving car hits a jaywalker, who is liable? The driver, the car manufacturer or the jaywalker?
The accident is caused by the negligence of the various parties involved. The pedestrian has violated the traffic rules and is partially at fault. By jaywalking, he is behaving irresponsibly and endangering the safety of himself, as well as other road users. Most self-driving cars today still require a human safety driver. In these self-driving cars, the control of the vehicle is shared between the driver and the car’s software designed by the software engineers. While the car is automated, the driver should remain alert and pay attention to traffic conditions even if the car had the right of way and be ready to take over in emergencies. Software engineers should also ensure that the car’s software and design can react to an emergency and make appropriate decisions before launching the product onto the market. The government could also exercise more caution and regulate autonomous vehicles by overseeing the safe testing and deployment of these emerging technologies. This can help to address safety requirements and safeguard the safety of road users to avoid future fatalities.
Also, there are ethical dilemmas on how the automated vehicles should be programmed as the machines will be required to make moral judgments and decisions when it encounters an unavoidable accident and the algorithm systematically favours or discriminates against a certain type of object to crash into. Consequently, the Moral Machine was deployed to gather human perspectives on the moral decisions made by self-driving cars. As discussed in lectures, the key elements of computational thinking are decomposition, pattern recognition, abstraction, algorithm design. The Moral Machine researchers used decomposition to break down the complex morality problem into smaller, more manageable parts and focus on different aspects when making a decision, such as saving more lives, protecting passengers, upholding the law, avoiding intervention, gender preference, species preference, age preference, fitness preference and social value preference. They then use pattern recognition to observe the patterns and trends and abstraction to focus on the moral preference of society. Algorithm design is used to develop instructions for the self-driving cars when encountering similar situations.
However, moral choices vary across cultures and are subjective and not universal. It involves judging the value of life. The consequence-based ethical theory focuses on the promotion of happiness and utility of the majority while the duty-based ethical theory focuses on the role of duty and respect for people. Based on the utilitarianism perspective, the car should minimise the number of casualties even if it requires sacrificing the people in the car. On the other hand, based on the deontological perspective, the car would prioritise minimising harm to the driver and passengers of the self-driving car. While a self-driving car that impartially minimises overall casualties would be more ethical, owners would likely prefer one that would protect passengers at all costs. This is an example of the classic tragedy of the commons.
The autonomous vehicle technology is still in its infancy stage and will continue to develop as technology advances. Besides the ethical concerns, there is also the risk of the self-driving car system being hacked. This could potentially lead to a series of collisions with serious consequences. Consequently, roads may be blocked and become inaccessible and the overall traffic flow will be affected. Hence, self-driving cars should also be designed to withstand malicious attacks. Despite the potential of self-driving cars, they still have a long way to go before they gain social acceptance.
