Artificial Intelligence vs. Human Intelligence: Who’s Driving Who?
- Selina Barker
- 4 days ago
- 4 min read
The Three Laws of Robotics are a popular talking point when it comes to AI. The first law in particular is often quoted or referenced in popular culture:
“A robot may not injure a human being or, through inaction, allow a human being to come to harm.”
The laws come from the sci-fi novels of Isaac Asimov, who used them as a plot device to ground his books. While causing no harm should certainly be the goal as AI advances, this is not always reflected in real-world outcomes, as evidenced in the recent push towards AI tech in cars.
New vehicle safety technologies can help improve road safety and reduce collisions, but so-called “self-driving” or “autonomous” features may give drivers a false sense of security. Automakers may not be as close to autonomous cars as they say.
Technically Safe
Advanced Driver Assistance Systems (ADAS), such as automatic emergency braking, lane-keep assist, and driver monitoring systems, are active crash-prevention technologies, designed to provide an extra level of protection by helping avoid preventable mistakes.
“90 per cent of road accidents are the result of human error.”
Many ADAS utilize Artificial Intelligence (AI) to recognize hazards. For example, some driver monitoring systems use in-cabin sensors that employ AI to detect a driver’s eye movement, head position, and even vital signs to identify signs of fatigue or distraction, like closed eyes or face rubbing.
Ultimately, vehicle safety features are meant to keep all road users safe from collisions, including drivers, passengers, pedestrians, and cyclists, and AI is helping to accomplish that. According to the Australian Academy of Science, “despite the increase of both people and vehicles on our roads, driving is safer today than at any time in the past.”
However, even the best technology can struggle, and even fail, when faced with complex and abnormal situations, and ADAS is no exception.
A Case for Accountability
In 2018, a “self-driving” car with a human “safety” driver behind the wheel was operating in autonomous mode during a test, when the vehicle struck and killed a pedestrian. Although the vehicle detected the pedestrian in time to stop, it did not determine that braking was needed until it was too late and did not stop or even slow before hitting the pedestrian.
This begs the question: who is accountable for this error in judgement? Is the human driver still responsible for braking when the vehicle is supposedly operating in autonomous mode, or is the vehicle’s computer accountable for failing to stop the vehicle after detecting a hazard? Can a computer ever be held accountable? The company that owned the self-driving vehicle was not held criminally responsible. However, they and several other companies did pull their self-driving test vehicles off public roads.
This is the first known case of a pedestrian being killed by a so-called “self-driving” vehicle and is a study of the pitfalls of relying on tech like AI behind the wheel. In today’s AI-forward market, many vehicle manufacturers are marketing "self-driving” to buyers and touting AI-powered features that make driving more convenient. But AI is not infallible, and while safety technologies can help mitigate drivers’ risk of collision, they cannot eliminate it.
“...vehicles have been involved in many collisions as a result of drivers taking the “self-driving” marketing literally.”
Becoming overly reliant on automotive AI can increase your risk of becoming complacent and disengaged behind the wheel. Even as in-vehicle safety technology continues to improve and automakers innovate new AI capabilities, the necessity to pay attention and remain alert and engaged cannot be overstated.
You Are Irreplaceable
The bottom line is that there are still no fully autonomous vehicles on the road. Automakers are operating on a 5-level scale of autonomy, but none are actually autonomous as of yet. Even semi-autonomous vehicles have hit plenty of roadblocks (no pun intended) and cannot be left unsupervised, needing a driver to monitor things either from the driver’s seat or remotely.
AI is not yet able to take over for drivers. ADAS can help reduce your risk of collision, but there is no replacement for being alert, being careful, and improving your defensive driving skills. One small way to reduce your risk of collision is to familiarize yourself with the ADAS in your vehicle and take advantage of them according to the manufacturer’s intended use. To go a step further in your safety journey, you can start to increase your safety knowledge and defensive driving skills by practicing techniques like blind spot monitoring, maintaining a following distance according to the weather, and gauging your vehicle’s limitations.
All of these techniques and more are part of Alertdriving’s comprehensive defensive driver training catalogue. With thousands of lesson topics that cover top collision causes, we routinely see our driver training reduce fleets’ incident rates – and we can do the same for you. Browse our Products page to learn more.
Sources:
Comments