Tesla robotaxi façade begins to crack, statistics still favour flesh and blood
For years, Elon Musk painted a future in which the human driver is as outdated and dangerous as an open flame at a petrol station. Fresh data now suggests that this vision rests on shakier ground than his confident promises imply. Tesla’s much heralded robotaxi dream, supposedly many times safer than an average person at the wheel, struggles to survive its first serious encounter with statistics. Under certain conditions, the company’s Full Self Driving system appears not just imperfect, but more dangerous than an occasionally distracted and tired human driver.
Instead of celebrating a breakthrough against road deaths, Tesla finds itself defending analyses that expose structural weaknesses in the system. Where a human instinctively reads complex situations and the body language of other road users, Tesla’s algorithms tend to freeze or perform unpredictable manoeuvres at precisely the wrong moment. According to incident data, cars operating in autonomous mode are involved in crashes more often than one would expect from Musk’s posts on X. This is no longer a matter of early software roughness. It is a growing crisis of trust in a system marketed as theoretically infallible.
Cameras versus reality
The shortcomings are most visible in city traffic and difficult weather conditions, where Tesla’s camera only philosophy begins to unravel. Unlike rivals that rely on expensive lidar and radar systems, Tesla insists that eight cameras are enough. The result is a series of situations in which the car fails to recognise roadworks, emergency vehicles, or simply unusual behaviour by other road users. Collisions follow that an average driver would likely have avoided. The emperor has no clothes moment cuts especially deep when Tesla raises prices while promising that every car can double as a money making robotaxi while its owner sleeps.
Regulators have also grown less tolerant of marketing exaggeration. Investigations suggest that Tesla’s safety statistics leaned heavily on ideal conditions, clear roads and motorway driving that resembled a train journey more than real world traffic. Once narrow streets and unpredictable pedestrians enter the picture, the house of cards collapses. The reminder is uncomfortable but familiar. Beta testing a new app is one thing. Experimenting with a two tonne machine on public roads is something else entirely.
Slow progress looks wiser
German and Japanese carmakers, advancing towards autonomy at a cautious and conservative pace, can afford a quiet sigh of relief. Tesla’s robotaxi vision currently looks less like a transport revolution and more like a clever financial narrative designed to soothe investors.
Humans make mistakes, but they tend to be predictable ones. Tesla’s software surprises in ways that leave no room for excuses or second chances. For now, flesh and blood behind the wheel remains more reliable than lines of code that still struggle to tell the difference between a plastic bag blowing across the road and a child running into it.
