If I could as Elon Musk One Question?

If I could as Elon Musk One Question? 

They say there is no such thing as a stupid question and If I could ask Elon Musk one question, it would be:

"How in the future will Tesla vehicles be able to drive safely using full unsupervised autonomy using just cameras? How for instance could a Tesla vehicle navigate safely on a road at night with really thick fog?"

Over the years, there has been a lot of speculation about the validity of Elon Musk's vision to make Tesla automobiles that are genuinely autonomous—meaning cars that can drive themselves without any human supervision. I have spoken to so many people who say Elon's vision is an impossible pipe dream. 

I have NEVER doubted Elon's vision, but there were two important variables I questioned.

The dream of autonomous transportation, as it was originally put forth, was to have a Tesla vehicle that could drive itself from San Francisco to New York, without a person in the vehicle. 

When Tesla first introduced FSD (Full Self Driving), there were two questions that would not stop popping into my head.

The first question, which I considered was "If a Tesla was going to be able to drive itself from San Francisco to New York, by itself, how would it charge itself along the way? 9 years ago, Tesla showed off this bizarre video that shows a prototype snake charger, that could robotically charge a Tesla through the standard wall charger outlet:

I always thought the solution above was NOT realistic or viable, but at the time I was already aware of wireless charing solutions for Electric Vehicles as I had attended the CES show in Las Vegas, and saw a working prototype that was similar to a wireless charger for a smartphone.

Back in 2020 I wrote a seminal article titled "Thoughts on the Future of Autonomous Vehicles: The Shape of Things To Come", in which I discussed my ideas on the future of autonomous vehicles. In that article, I included a profound video that showcased and documented the future reality of wireless EV charing becoming standard for autonomous cars, which can be seen below:

Recently, when Tesla announced the CyberCab, they mentioned they were going to be incorporating wireless charging, and thus one of the two final puzzle pieces was put in place. Below is an official Tesla Video that shows how their CyberCab will use wireless charging:

Of course, I realized this years ago, which was further confirmed with the advent of the CyberTruck as Sandy Munro's CyberTruck teardown discovered wire harnessing already built-in to the CyberTruck, that could simply come to life if a Tesla wireless charing pad was installed. In other words, all CyberTrucks can easily be upgraded to have wireless charging. 

So this solves half the equation. Also, it's easy to see the amazing benefits of wireless charing for Tesla vehicles. For instance, your Tesla could pull itself into your garage automatically and position itself perfectly over the wireless charing pad. I have even seen prototypes where the base/floor charing pad was on wheels that could move horizontally, vertically, or diagonally so it could optimize its charing position every time. Jordan Giesige, from The Limiting Factor just came out with a deep dive video on how Tesla Wireless Charing will work, which is interesting:

How "Tesla AI Vision" Works

The second half of the equation I set out to resolve, is based on the Tesla exterior camera system, which the car used to drive when placed in full self driving mode, which as of today is only available in supervised mode.

In other words, Tesla and Elon Musk have promised that in the future, they plan to release Full Self Driving (Supervised), which will allow Tesla models to drive themselves.

Several years ago, all Tesla models had both cameras and lidar sensors for autonomous driving. The logic was that whatever the camera couldn't see, the lidar could detect. Lidar is a form of radar, that can see through things.

An edge case or corner case complex driving challenge example would be:

What if an autonomous Tesla was driving itself on a slippery road where snow was falling, and there was thick fog, and let's say there was a person crossing the road, or traffic stopped ahead of your, that you could not see since the fog is so thick. If the car was traveling at 60MPH, and the cameras could not see through the fog, the lidar would see what was behind the fog, which makes perfect sense.

Vision Based Approach

But then a few years ago, Elon Musk informed the world that Tesla was completely removing lidar and all forms of radar from Tesla vehicles, and planned to move forward with a camera only approach. This confused me to no end, and with my scenario, I couldn't understand how a Tesla could drive itself safely in such a scenario as I just mentioned with all the fog, without crashing or posing a significant safety hazard.

I kept doing research on this subject and asked a number of smart people and nobody seemed to know the answer, but I recently stumbled into a video that addressed and seems to have resolved this issue. The video below is from BestINTesla and is titled: "The GPU Giants Bet Big on xAI - The Dawn of a New Era in Energy Storage - Tesla Triumphs Over Unions". In this video, around 9:44:00 he discusses this topic in detail:

In the video above he points out that Chinese EV manufacturer, Li Auto CEO voiced his concern about Tesla's camera only approach. 

Ironically, Li Auto CEO's scenario was similar to mine, but according to the video the way the Tesla cameras work, is they can use all the visible light, including in the infrared spectrum, which we as humans can't see or detect with our eyes, and can see individual photons, which serve as solution of measurement similar to lidar. This still sounds a little far fetched to me, but I plan to investigate more and once I figure out more, I will update this article.

So there you have it. If I could ask Elon Musk only one question, this would defiantly be it. One thing that still confuses me and seems to contradict this notion or solution, is why do I get FSD pop ups that say FSD is degraded due to poor weather conditions? In other words, if I drive in rain, often times Tesla TSD will either refuse to turn on altogether, and give me a message that says something like "FSD not available at this time. Please try again later", or it will turn on the FSD, and put up a message on the Tesla screen that says something like, "FSD quality is currently degraded at this time, so please pay extra close attention".

What Does Grok 2 Have to say?

I decided to ask X.ai's Grok2 the question just to see how it would answer it. I figured since Elon Musk owns Tesla, Twitter/X.com, as well as Grok2, it should be able to separate the fact from the fiction.

Below is the exact question I asked Grok2:

"How in the future will Tesla vehicles be able to drive safely using full unsupervised autonomy using just cameras? How for instance could a Tesla vehicle navigate safely on a road at night with really thick fog?"

Below is Grok2's exact and thorough answer:

Achieving full unsupervised autonomy with just cameras poses significant challenges, particularly in adverse conditions like navigating at night with thick fog. Here's how Tesla might address these challenges:

Camera Technology and Software Improvements: Tesla relies on an advanced vision system called "Tesla Vision," which uses a suite of cameras to gather visual data. The key to unsupervised autonomy would involve enhancing the capabilities of these cameras through:

Advanced Image Processing: Utilizing neural networks and machine learning to interpret visual data more effectively. These algorithms can learn to recognize patterns in low visibility or altered conditions, potentially identifying road boundaries, obstacles, and other vehicles even when visibility is low.[](https://caradas.com/tesla-vision-adas-features/)

Night Vision Capabilities: Cameras with sensitivity to infrared light could improve night-time visibility, allowing the vehicle to detect objects that are not visible to the human eye. This would be crucial for scenarios involving dense fog, where traditional visible light cameras struggle.[](https://x.com/PhilFox43249797/status/1873667047336591851)

Redundancy and Fail-Safe Systems: Even with just cameras, Tesla could implement systems where multiple angles are captured and cross-referenced to compensate for any single camera's failure or obstruction. This could include predictive algorithms that anticipate road layout based on prior data and mapping.[](https://www.tesla.com/support/autopilot)

Challenges and Limitations:

Fog and Weather Conditions: Thick fog can obscure all types of visual sensors, including cameras. Tesla's approach would need to significantly improve the cameras' ability to penetrate fog or develop software that can infer the environment from partial or intermittently available data. Currently, cameras alone are at a disadvantage in such conditions compared to systems using lidar or radar, which can better handle low visibility.[](https://caradas.com/tesla-vision-adas-features/)[](https://x.com/inevitable360/status/1873664929200538033)

Hardware Limitations: The current camera setup on Tesla vehicles has been criticized for not providing the depth perception and redundancy needed for true Level 5 autonomy, especially in scenarios where cameras can be easily obstructed or fail due to environmental factors.[](https://www.reddit.com/r/SelfDrivingCars/comments/1f7x2dr/your_tesla_will_not_selfdrive_unsupervised/)[](https://www.theglobeandmail.com/drive/mobility/article-all-eyes-on-teslas-cameras-going-it-alone-on-driver-autonomy/)

Future Prospects:

- Tesla's ongoing software updates and data collection from its fleet could lead to breakthroughs in how vehicles interpret visual data under various conditions. Continuous learning from real-world driving scenarios could improve the system's performance in fog, night, and other adverse conditions over time. However, experts and posts on X remain skeptical about achieving full unsupervised autonomy with cameras alone due to these inherent limitations.[](https://x.com/TheBenHakim/status/1875264250631090400)[](https://x.com/TheBenHakim/status/1874927342373720475)

In summary, while Tesla is pushing forward with a camera-only approach, the journey to full unsupervised autonomy in challenging conditions like thick fog at night involves overcoming significant technical hurdles. It would require advancements in camera technology, machine learning algorithms, and possibly some level of sensor fusion or environmental adaptation that hasn't yet been fully realized or demonstrated at scale.

My Summary of Grok2's Answer

I must admit I'm disappointed and relieved at the same time by Grok AI's answer. I'm disappointed because its answer was vague and unhelpful and didn't go as far as the answer I provided. Basically, I would say Grok2's answer was largely useless. I'm relived because now I don't feel like my question is stupid, if Grok can't even answer it.

TMAXX Shamrock Greet Matte PPF

Telsa Model X Plaid

TMAXX Shamrock Greet Matte PPF

I really like this brand new PPF color from TMAXX, which TSportline wrapped this Model X Plaid with!!! I remember a friend of mine, back in the late 1990s or early 2000s drove a BMW 7 series that had a really similar shade of forest green that was so beautiful, but it was not matte like this new PPF, which gives it a frosty cool look.