In a watershed moment for the automotive industry and artificial intelligence, a Tesla owner has successfully completed what has long been considered the "holy grail" of consumer autonomous driving: a full coast-to-coast journey across the United States with absolutely zero human interventions. Utilizing Tesla's Full Self-Driving (FSD) Supervised system, the vehicle navigated complex highways, city streets, parking lots, and charging stations entirely on its own, marking the fulfillment of a promise made nearly a decade ago.
The feat, accomplished by Tesla owner David Moss, utilized a Model 3 equipped with the company's latest AI4 hardware and FSD software version 14.2.1.25. The journey spanned 2,732.4 miles from Los Angeles, California, to Myrtle Beach, South Carolina, completed in just under three days. This milestone has sent ripples through the electric vehicle community and garnered immediate praise from high-ranking Tesla executives, signaling a potential paradigm shift in the readiness of autonomous technology for mass adoption.
The Journey: From the Tesla Diner to Myrtle Beach
The historic drive commenced at the Tesla Diner in Los Angeles, a symbolic starting point representing the brand's lifestyle integration. David Moss, piloting a stealth gray Model 3, set his navigation for the East Coast. According to the data released following the trip, the entire duration of the drive was 2 days and 20 hours. While cross-country cannonball runs are a staple of automotive lore, this specific journey is unique in history because the driver reportedly never touched the steering wheel or pedals to correct the vehicle's trajectory.
Moss documented the experience on the social media platform X (formerly Twitter), providing real-time updates and subsequent data verification. "I am proud to announce that I have successfully completed the world's first USA coast to coast fully autonomous drive," Moss wrote in a celebratory post. He detailed that the trip concluded in Myrtle Beach, South Carolina, covering a staggering 2,732.4 miles of diverse American roadways.
Perhaps the most critical aspect of this achievement is the definition of "zero interventions." In the context of autonomous driving testing, an intervention or disengagement occurs when the human driver must take control of the vehicle due to safety concerns, software confusion, or mapping errors. Moss stated unequivocally, "This was accomplished with Tesla FSD V14.2 with absolutely 0 disengagements of any kind even for all parking including at Tesla Superchargers."
This claim implies that the vehicle not only handled high-speed interstate travel but also managed the complex logistics of exiting highways, navigating to Supercharger stalls, backing into charging spots, and re-entering traffic flows—all without human assistance. Moss further clarified in later comments that there were "zero close calls" during the trip, suggesting a smooth and confident performance by the AI.
Technological Leap: AI4 and FSD v14.2
The success of this coast-to-coast drive highlights the rapid capabilities of Tesla's hardware and software integration. Moss's vehicle is equipped with Hardware 4 (AI4), Tesla's latest suite of cameras and processing computers designed to offer higher fidelity and faster processing speeds than the previous Hardware 3 iteration. While Tesla has maintained that HW3 is capable of full autonomy, the seamless nature of this trip on AI4 suggests that the newer hardware provides a significant reliability buffer.
The software at the heart of this achievement is FSD v14.2.1.25. This version represents a mature iteration of Tesla's "end-to-end" neural network approach. Unlike earlier versions of Autopilot that relied on hard-coded rules for driving behavior (if x, then y), modern FSD versions utilize neural networks trained on millions of miles of video data to make driving decisions. This mimics human learning, allowing the car to understand context and nuance in driving scenarios that rule-based code simply cannot handle.
According to community trackers and the FSD database, Moss had driven 10,638.8 miles in his Model 3 prior to this specific trip, with the system handling 100% of the driving during the recorded sessions. The specific software version, v14.2.1.25, had been installed on his vehicle only 12 days prior to the completion of the trip, indicating a rapid adaptability of the software to real-world conditions immediately upon release.
Verifying the Milestone
In the world of autonomous driving claims, verification is paramount. Skepticism often accompanies reports of "zero intervention" drives due to the subjective nature of what constitutes a necessary intervention. To address this, Moss directed interested parties to third-party verification tools.
"You can verify all the data using @wholemars FSD database," Moss posted, linking to the community-driven tracker that logs telemetry data from Tesla vehicles.
The utilization of the FSD database allows the community to scrutinize the drive logs, ensuring that the autopilot system remained engaged throughout the duration of the trip. This transparency is crucial for building public trust in the technology. The data confirms the route, the engagement status of the FSD system, and the lack of disengagements, solidifying the legitimacy of the record.
Fulfilling a Decade-Old Prophecy
To understand the magnitude of this achievement, one must look back to October 2016. When Tesla unveiled Autopilot 2.0, CEO Elon Musk made a bold prediction: a Tesla would be able to drive itself from Los Angeles to New York City without a single touch from the driver by the end of 2017. A promotional video released at the time showed a Tesla navigating local roads and highways, set to the Rolling Stones' "Paint It, Black," but critics and engineers alike noted that the technology was not yet ready for a dynamic, unscripted cross-country run.
The year 2017 came and went, as did several subsequent years, with the "coast-to-coast" drive becoming a recurring question in earnings calls and shareholder meetings. The delay gave rise to the term "Elon Time," referencing the CEO's tendency to set overly optimistic timelines for technological breakthroughs. However, despite the delays, the objective remained a core target for the Autopilot and AI teams.
With David Moss's successful drive in late 2025, the prophecy has finally been fulfilled. While it arrived significantly later than the 2017 estimate, the accomplishment vindicates the vision that a camera-based, neural-network-driven approach could solve the complexities of general autonomy. As the Teslarati report noted, "It appears that Elon Musk's estimate was not impossible at all. It was just late."
Executive and Community Reaction
The reaction from Tesla's leadership was swift and celebratory, underscoring the internal significance of this external milestone. Elon Musk acknowledged the feat on X, joining a chorus of congratulations from the Tesla community. However, the most notable validation came from Ashok Elluswamy, Tesla's Vice President of AI Software.
Elluswamy, who has been instrumental in the development of the end-to-end neural network architecture, posted:
"World's first fully autonomous coast-to-coast drive, done with Tesla self-driving v14. Congrats and thank you @DavidMoss!"
The official Tesla North America account also amplified the news, stating, "First Tesla to drive itself from coast to coast w/ FSD Supervised. 0 interventions, all FSD." This official acknowledgement serves as a corporate stamp of approval on the user-generated achievement, signaling that Tesla views this not just as an anecdote, but as a demonstration of product maturity.
The Implications of "Supervised" Autonomy
It is important to note that the system used is still classified as "FSD Supervised." This distinction remains critical from a legal and safety standpoint. Unlike Waymo or Cruise vehicles, which operate as Level 4 autonomous robotaxis in geofenced areas without a driver, Tesla's FSD currently requires a licensed driver to sit in the driver's seat and remain attentive, ready to take over at any moment.
David Moss's drive, while requiring zero physical interventions, still involved 100% human supervision. Moss was present to monitor the vehicle, ensuring it adhered to laws and acted safely. However, the fact that he did not need to intervene over nearly 3,000 miles suggests that the gap between "Supervised" and "Unsupervised" is closing rapidly. The data generated from such trips is invaluable to Tesla's goal of proving to regulators that the system is statistically safer than a human driver, a prerequisite for removing the supervision requirement.
Navigating the Complexity of American Roads
A coast-to-coast drive is the ultimate stress test for an autonomous system because of the sheer variety of driving conditions encountered. A route from Los Angeles to South Carolina traverses the dense, aggressive traffic of Southern California, the open deserts of Arizona and New Mexico, the variable weather patterns of the South, and the distinct road markings and traffic laws of over half a dozen states.
The AI had to contend with:
- Construction Zones: Variable lane shifts, cones, and temporary signage that often confuse standard sensors.
- Weather Conditions: Differing lighting, potential rain, and glare across different time zones.
- Charging Logistics: Navigating into Supercharger stations, which are often located in busy parking lots behind shopping centers, requires precise low-speed maneuvering and interaction with pedestrians.
- Fatigue Management: While the car doesn't get tired, the system must maintain peak processing performance for hours on end.
That FSD v14.2 handled all these variables without a single disengagement speaks to the robustness of the AI's generalized learning. It suggests that the system is no longer memorizing maps but is genuinely "seeing" and understanding the world in real-time.
The Future: From Feature to Robotaxi
This milestone serves as a foundational proof-of-concept for Tesla's upcoming Robotaxi ambitions. For a dedicated Robotaxi network to function, vehicles must be able to operate for extended periods without human input. A 3,000-mile zero-intervention streak is effectively a simulation of a Robotaxi operating for weeks without needing remote assistance or rescue.
If fleet data begins to show that zero-intervention drives are becoming the norm rather than the exception, the argument for removing the steering wheel—as seen in Tesla's Cybercab concepts—becomes much stronger. The industry is watching closely to see if Moss's experience can be replicated by other drivers in different weather conditions and on different routes.
Conclusion
David Moss's coast-to-coast journey is more than just a road trip; it is a historic data point in the timeline of transportation. It represents the transition of autonomous driving from a futuristic experiment to a tangible reality capable of handling the breadth of the American landscape. While the timeline was longer than originally predicted in 2016, the result is no less impressive.
As Tesla continues to refine FSD v14 and push updates to its fleet, the frequency of such zero-intervention trips is likely to increase. For now, David Moss holds the title for the first verified autonomous cross-country drive, proving that the vision of a car driving itself from the Pacific to the Atlantic is no longer science fiction—it is a feature available in the driveway today.
