When Will Self-Driving Cars Be Truly Safe?

The promise of self-driving cars is undeniable, with many believing they can drastically reduce traffic accidents and fatalities. However, the question remains: when will these vehicles truly be safe?

The notion that self-driving cars need to be 'as safe as human drivers' is a misnomer. Most human drivers break numerous traffic laws, speed, drive aggressively, and cause accidents. Calling this 'safe' is a joke. Self-driving cars only need to be better than the average driver, not the best. But before we can consider this, they need to prove they can match top human drivers in real-world conditions.

Demonstrating Superior Driving Capabilities

To achieve this, self-driving cars should undergo rigorous testing under high-pressure, competitive conditions. Let's take NASCAR and Formula 1 races on road courses as examples. Success in such environments might suggest they can handle complex road conditions and high-stakes situations with precision. Only then can we seriously consider self-driving cars as a viable alternative to human-driven vehicles.

Currently, even though self-driving cars may not reach the same level of skill as the best human drivers, they operate with fewer dangers. For instance, they don't get tired, distracted, or engage in activities like eating, drinking, texting while driving, or driving under the influence of alcohol or drugs. These factors make them inherently safer than human drivers.

However, this doesn't mean they are perfect. Human drivers often drive despite adverse weather conditions, knowing they are poor judges of risk. Self-driving cars, on the other hand, might avoid driving during such conditions, which could be seen as avoiding normal traffic situations where they would gain real-world testing experience.

The Current State of Self-Driving Cars

Many companies have claimed their self-driving technologies are ready, with final tests underway. Yet, complete safety cannot be guaranteed unless these vehicles are tested in real-life scenarios over a substantial period. It is essentially impossible to declare a system safe without extensive public use and real-world data.

The absence of clear regulations and benchmarks means there is no defined framework for determining safety standards. If there is a committee defining rules, it will likely bring structure to the process, but comprehensive safety standards are still a work in progress.

Defining 'Safe'

The concept of 'safe' can vary widely. If 'safe' means being no more dangerous than human drivers under the same conditions, then we may already be there. Modern self-driving technologies excel in specific areas where they have extensive data, such as Waymo's limited suburban areas. Tesla, while less capable in these areas, covers a broader range and is still more competent than human drivers.

To achieve true safety, it may be more appropriate to compare the management style of these systems. For example, Waymo operates like a micro-manager, ensuring every detail is perfect. In contrast, Tesla delegates more authority, allowing for more flexibility and adaptation to different road conditions.

In conclusion, while the potential of self-driving cars is immense, achieving true safety requires rigorous testing and real-world use. The journey towards safer driving is ongoing, but the advancements made so far are promising.