In a recent and ambitious forecast that has sent ripples across the motor racing and car-making world, Elon Musk has conjectured that version 14 of the forthcoming Full Self-Driving (FSD) system would be definitively superior to human performance, with possible improvements up to 2- 3x more powerful than human capabilities. He also hinted at a 10-fold increase in the capability of version 15.
What’s New in FSD V14?
Tesla V14 is expected to become one of the largest updates to AI/Autopilot in the company’s history. V14 will increase the parameter count by 10x, according to Musk, and thereby take into account and analyze significantly more real-world data peculiarities. Such a boost in AI processing capability should provide more nuance, contributing to context-sensitive decision-making within the traffic. What’s interesting is that even with the current FSD V13.2.6, Tesla managed to tackle China’s urban maze with not even a scratch.
He also stressed that the system will greatly decrease the frequency with which it has to nag the drivers for attention, so we can hope that the experience will be far less intrusive and hectic, with the necessary degree of oversight being present.
How Much the Better? Musk’s Metric
According to a recent translation of Musk’s remarks, “Tesla FSD V14 will definitely be better than human driving, but I am unsure if it will achieve a level that is ‘ten times better than humans’; it may be two to three times better. The V15 version could reach ten times.” Though a third party translates the sentence, it is aligned with the message that was conveyed by Musk – there will be an improvement in V14, and V15 is even higher.
Roadmap and Expectations
Tesla enthusiasts await with hope to see V14 soon. Musk has earlier suggested that a breakthrough in driving AI is on the horizon, with a 10x increase in parameters and an even sharper decline in the number of driver requests, which are about six weeks away. Assuming that everything went according to plan, they would have some early access in the range of late September, with the rest of the world coming at around the holiday season.
Culture of Skepticism and Past Performance
Even though these projections are very daring, Tesla has been notorious whenever it comes to delivering on Musk’s vision in terms of time schedules. A post by Wired identifies a pattern of hyped autonomy declarations—robotaxis, hyperloop, humanoid robots—that over and over fail to live up to their billing.
Up to this point, FSD V13.2 has demonstrated significant improvements. As denoted by the Toughest Real-World Driving Test by Dongchedi in China, by employing a closed public highway, the team tested 36 vehicles under six complex and dangerous scenarios. Nevertheless, the gap between successive refinements and a system that consistently fares better than humans behind the wheel is very wide indeed.
Provided that Tesla does not fail to actualize its promises, FSD V14 promises to reshape the notion of assisted driving as perceived by the people. A future system that is provably or measurably safer and more human-like may drive acceptance more rapidly, affect insurance models, and change the regulatory attitude toward autonomy.
At least, until Musk succeeds in driving the vehicle in the real world under the real circumstances, this ambitious vision, when it says 2 to 3 times better than humans today and 10 times better by next year, has yet to be proved.
