Cartorque Vol 19 - ‘Rise of the Robotaxi’
As I’m sure you’ve heard, driverless cars are actually a thing now. This entry into our catalogue of waffling was planned for earlier in the year but we never expected the exponential growth of such a cyberpunky idea. The first location was in Pheonix, chosen because of the wide predictable road patterns and predictable weather. From there, San Francisco was next, but with it taking four years to go from one city to two it was easy to be sceptical and write it off as an engineering project. However, as you can now get driven by a robot in Dallas, Houston, San Antonio and Orlando and plans of expanding to our backyard by summer 2026. We feel it was time to give our two cents.
Unsurprisingly, it is Alphabet (Google’s parent company) spearheading the movement of eliminating the driver through Waymo. Beginning the endeavour in 2009, I am sure we all remember their early efforts of mounting a spinning laser to the roof of a Prius or something. 16 years later and the core idea remains the same. Mounting a suite of sensors and radar to a Jaguar I-Pace and plonking an AI behind the controls.
“We all know how reliable those are” is something I can already feel you thinking. I counter with reports of up to a 92% reduction of accidents compared to average human drivers. Unlike humans, an AI will follow rules it is told to the absolute letter which means that your soon-to-be local driving robot will most likely be a better driver than you are. Unfortunately though, an AI will follow rules it is told to the absolute letter which means that someone wearing a stop sign on a t-shirt has the ability to stop a robotaxi (yes, this happened) or hard braking for life sized posters (yep, this one too).
Their inherent inflexibility is a problem, one that especially is worrisome for a tight city like their proposed plans for London. A person can gesture to other drivers to work out problems, two cars working their way down a narrow street for example. A driverless AI instead might decide “scenario unclear, I shall remain here indefinitely” as it wasn’t in its training data. Leading to an obedient robotaxi politely blocking traffic while trying to understand a traffic cone.
Though this hyperfocus on objects is annoying to everyone, including the passenger, it is incredibly safe. It may mistake a plastic bag for an incoming asteroid but at least it isn’t checking Instagram at the lights. It won’t get tired or bored, need to take shifts or must entertain the inevitable “been busy tonight?” conversation.
And thus, we are at an impasse, the traditional taxi has been mostly overtaken by Uber/app-to-order. Reducing the human element to just the in-car experience. Now could we see Uber, or the app-to-order model, lose out to the driverless equivalent? It’s possible but this tech is still in its infancy. Resistance will exist, as it does with any technological advancement and it is extremely sensitive. In the unlikely event we wake up tomorrow and see a serious crash caused by a driverless car, the whole project will be scrapped and shunned by governments worldwide. In the more likely event that we will continue to see robotaxi’s having horn honking competitions in car parks, how they will integrate into the economy regarding jobs becomes the more pressing issue.
To round it off, the robotaxis are coming. Waymo is the frontrunner, Amazon and Tesla have their own projects in the works. With promises of fewer crashes, less social awkwardness and the ability to get you to your destination without judgement they certainly have their appeal. However, the little AI chauffeurs still have much to learn as they navigate the world’s roads. Still, they are showing remarkable potential. A political and economic minefield looms, but interesting times are ahead in the world of taxis.
Writer
Sean Isaac
