01-03-2023 | By Robin Mitchell
Recently, Tesla CEO Elon Musk has been slapped with another lawsuit from shareholders, claiming that promises surrounding Autopilot were false, resulting in a fall in share price. The lawsuit comes after a series of failed attempts to deliver on Musk's self-driving promises, which have caused stock prices to fluctuate. Despite setbacks with Tesla, Musk continues to innovate in other areas, such as SpaceX's rocket launches and EV development, but the recent lawsuit raises questions about the future of his leadership. Why is self-driving such a complex problem to solve, what does the lawsuit mean for Elon Musk, and could this be the start of the end for Elon Musk and his false promises?
Despite numerous attempts to develop self-driving vehicles, they are yet to be publicly deployed. Vehicles that are capable of steering themselves and adjusting their speed may be considered by some as self-driving, but in reality, these features are over-glorified adaptive cruise control. Some users have been able to go on entire journeys without needing to take control of their vehicles, but while these vehicles could be considered self-driving, these journeys are either carefully planned on specific roads or not subjected to unexpected changes such as road works. At the same time, vehicles with this capability require users to hold the wheel and be alert at all times, meaning the vehicle is not allowed to drive autonomously, which a self-driving vehicle can do.
But why is self-driving such a monumentally difficult task to complete? It turns out that what humans perceive as simple is actually computationally difficult for digital machines. The human brain can receive millions of signals every second, decoding the data to understand what is happening and then make decisions based on that data. At the same time, the human brain can converse with someone else, deciding what to have for dinner, what film to watch later, and recall past events.
However, ask a human to multiply two large numbers, and it can take a long time to get a result. On the flip side, digital circuits are able to process massive amounts of serialised data and perform complex mathematical operations, all while using an extremely small amount of energy. But ask a computer to analyse a single 4MP image and identify objects, and the result is extremely slow. Using multiple processors in parallel can help with such tasks; however, this comes at a high economic and energy cost.
Thus, trying to create a self-driving vehicle with multiple cameras and sensors facing all directions introduces serious data challenges, and modern computational devices are still unable to process the same amount of information that the human brain does when driving. This data load can be reduced by using two front-facing cameras only, but this introduces the possibility of missing key details, such as vehicles at the rear and sides.
Computers used to control vehicles would also need to be able to reliably identify objects in the road, whether it is a pedestrian, obstacle, police, or roadworks. The high degree of variability in a journey and what can be encountered means that training an AI to recognise every possible scenario is impossible. Overall, self-driving is a computational challenge that modern computing still struggles to deal with.
Elon Musk is well known for making big claims, numerous promises, and guarantees, whether it’s his aim to colonise Mars by 2020, boring tunnels under LA, creating a hyperloop that will cost next to nothing for passengers, or implanting chips into people’s brains. One promise that he has repeatedly made for several years now is that Tesla Full-Self Driving will be ready by next year, every year. And, of course, as each year passes, this feature is still yet to be officially launched on all Tesla vehicles.
However, while the Tesla fanboys continue to support Elon through his ambitious plans, some Tesla shareholders have had enough and recently launched a lawsuit against Elon Musk on the grounds of false promises with the autopilot technology and the resulting fall in shares. Specifically, the shareholders have expressed anger at the claims that have been made concerning the safety of Tesla’s self-driving feature. Numerous fatalities have occurred as a result of Tesla vehicles, primarily due to the use of self-driving software, and while traffic fatalities are a daily occurrence, data is starting to emerge that Tesla vehicles may be presenting an increased safety risk to drivers.
When safety concerns surrounding Tesla’s self-driving features were brought to the National Highway Traffic Safety Administration and the Securities Exchange, Tesla share prices took a substantial hit. A further drop in value followed after 362,000 vehicles had to be recalled due to Full-Self Driving having issues with intersections.
“As a result of defendants’ wrongful acts and omissions, and the precipitous decline in the market value of the company’s common stock, plaintiff and other class members have suffered significant losses and damages,” – Extract from the complaint, led by shareholder Thomas Lamontagne
Many in the engineering community understand the oddity that is Elon Musk and how most of his promises are empty. Generally speaking, Elon Musk will recycle an ancient idea, present it as original, raise large sums of money to fund it from outsiders, and then let the project die slowly as the public loses interest. For example, the solar panel’s roof tiles have yet to become available to consumers, Tesla vehicles still can’t drive themselves, no chips have been implanted into human brains, and the Hyperloop has been reduced to a rusty length of tube in the desert. Even the SpaceX Starship is yet to take off and prove itself, despite numerous promises of its use with Mars missions three years ago.
Now, Elon Musk faces multiple lawsuits from most of his companies, which could be the beginning of the end for Elon Musk. Many will believe that he is a visionary who wants to advance the human race, and this could very well be true. However, if Elon Musk doesn’t start delivering results with his EVs, rockets, and other out-there technologies, the world may begin to lose faith in his ability to develop new technologies.