Tesla's Full Self Driving V12 is showing promising improvements in human-like behavior and performance, with the potential for significant advancements in self-driving technology
Questions to inspire discussion
-
What improvements does Tesla's Full Self Driving V12 show?
—Tesla's Full Self Driving V12 shows promising improvements in human-like behavior, roundabouts, and overall stability, with positive feedback from testers.
-
How are Tesla employees testing Full Self Driving V12?
—Tesla employees are testing Full Self Driving V12 internally, with the update being rolled out in waves to executives, team leads, and eventually the Early Access program for employee-owned vehicles.
-
What is the feedback on Full Self Driving V12 from Tesla employees?
—Feedback from Tesla employees on Full Self Driving V12 indicates improvements in human-like behavior, stability, roundabouts, and lane keeping, with a majority reporting no change or better performance.
-
What is the potential release date for Full Self Driving V12?
—Tesla employees are hopeful for the release of Full Self Driving V12 in late January, with potential delays due to the focus on finalizing a stable build for the Cybertruck and smart summon shipping in the later end of q1.
-
How is Tesla using end-to-end neural nets to improve their full self-driving system?
—Tesla is using end-to-end neural nets to improve the full self-driving system, making it easier to implement modifications and customization for different user preferences based on fleet data and human feedback.
Key Insights
- 🚗 Many Tesla investors are eagerly awaiting the wider release of full self-driving V12, expecting a significant step change in the technology.
- 📈 "The fact that they're testing it with V12 is a good indication that V12 can finally solve a lot of the issues that were difficult to solve in the previous regime."
- 🚗 72% of Tesla employees reported improvements in human-like behavior and general stability with Full Self Driving V12.1.
- 🤖 The Full Self Driving V12 is all neural net, it didn't code anything, it just learned from videos and behaves more human-like in driving situations.
- 🚗 70% of Tesla employees felt that Full Self Driving V12 was actually acting as it should, showing positive feedback on its performance.
- 🔄 The self-learning capability of Full Self Driving V12 means that the more it gets out there, the more it learns and improves.
- 🤐 The FSD beta tester's feedback on Full Self Driving V12 is highly respected and critical, but their name is kept anonymous.
- 👍 It's important to have feedback from experienced testers who are forthcoming about both the strengths and weaknesses of the system.
#Investments #Tesla #FSD @herbertong @HansCNelson
Clips
-
00:00 🚗 Tesla employees provide feedback on Full Self Driving V12, highlighting improvements in human-like behavior and roundabouts, but express uncertainty about its overall significance and the need for testing in different environments.
- Tesla investors are eagerly awaiting the release of Tesla's full self-driving version 12, and feedback from employees suggests improvements in human-like behavior and roundabouts.
- Tesla employees are unsure if the Full Self Driving V12 update will be a significant improvement, as it has only been tested in California and needs to be tested in different environments.
- Tesla has begun internal testing of smart summon on company-owned vehicles based on version 12's architecture, signaling good progress and potential solutions to previous issues.
- Version 12 of Full Self Driving was rolled out to Tesla employees in waves, starting with executives and team leads, and eventually expanding to the Early Access program for employee-owned vehicles.
- Tesla employees discuss the potential features of Full Self Driving V12, including park assist visualization and the ability for the car to identify and park in viable spaces.
-
05:25 🚗 72% of Tesla employees see improvements in Full Self Driving V12, with the system needing predictable behavior for human driver engagement, using a neural net to make driving more humanlike and safe.
- Tesla employees feedback on Full Self Driving V12 shows 72% expressing improvements in human-like behavior and general stability, 20% noting no change, and 8% experiencing some worse than before.
- The full self-driving system needs to have predictable and comfortable behavior for the human driver to remain engaged, even at level four, in order to anticipate and expect traffic.
- The new Tesla Full Self Driving V12 uses a neural net that learns from videos to make the driving experience more humanlike and improve safety.
-
08:27 🚗 70% of Tesla employees felt Full Self Driving V12 was acting well, with improvements in humanlike behavior, roundabouts, and lane keeping, and Elon Musk emphasized the need to test it in different weather conditions.
- 70% of Tesla employees felt that Full Self Driving V12 was acting well, with 40% expressing improvements with roundabouts and 50% noting no change, while 10% noted a worse experience, but overall, 90% reported no change or better performance.
- Tesla employees provided feedback on Full Self Driving V12, with 72% expressing improvements in humanlike behavior, 40% in roundabouts, and 37% in lane keeping, while a majority reported no change and very few reported a worse experience.
- Tesla employees in California are testing Full Self Driving V12 and Elon Musk emphasized the need to test it in different weather conditions.
-
11:48 🚗 Full Self Driving V12 needs ongoing learning to adapt to different road conditions and environments, utilizing a large fleet and data to improve performance and reach impeccable system safety for autonomous driving.
- Learning to drive in different environments and weather conditions is similar to how Full Self Driving V12 needs to adapt to different cities and their driving etiquette and surroundings.
- The FSD driving system needs ongoing learning to adapt to different road conditions and environments, utilizing the large fleet and data to improve performance in various geographies.
- The system needs to be overfit for safety and performance, but as progress is made, they will need to solve driving variations on their own and constantly improve the overall architecture to reach impeccable system safety for autonomous driving.
-
15:24 🚗 Tesla employees report significant improvement and self-learning capabilities in Full Self Driving V12, with a 3.5 times rate improvement compared to V11, hopeful for late January release but potential delays due to Cybertruck focus and smart summon shipping in q1.
- Feedback from 15,000 Tesla employees using Full Self Driving V12 indicates significant improvement and self-learning capabilities, with a reported 3.5 times rate improvement compared to version 11.
- Tesla employees are hopeful for the release of Full Self Driving V12 in late January, with potential delays due to the focus on finalizing a stable build for the Cybertruck and the likelihood of smart summon shipping in the later end of q1.
- The release of V12 for Full Self Driving may be delayed due to extensive testing needed outside of California, but once deployed to customers, progress should accelerate.
-
18:48 🚗 Tesla employees and experienced testers are impressed with the improvements in Full Self Driving V12, despite some critical feedback on its shortcomings.
- An experienced FSD beta tester went for a ride with a Tesla employee to test out FSD beta 1 and shared their critical feedback on the shortcomings of the system.
- Tesla employees are impressed with the Full Self Driving V12's performance during test drives, with only a few instances of disengagement in three years.
- Version 12 of Tesla's Full Self Driving showed improvements in merging, speed bump detection, and natural slowing to stop, with positive feedback from experienced testers.
-
21:37 🚗 Tesla is expanding Full Self Driving into customer hands, using end-to-end neural nets to improve the system and actively engaging with feedback to make modifications and customization easier.
- Tesla is expanding Full Self Driving into customer hands, with rumors of a launch in China for version 11 or possibly version 12.
- An end-to-end solution for full self-driving would be easier to adapt to different areas of the world than a hand-coded approach, and a V12 based version would be what they want to release.
- Tesla is using end-to-end neural nets to improve the automatic wipers based on fleet data and human feedback.
- Tesla employees are actively engaging with feedback to improve the full self-driving system, with the end-to-end architecture making it easier to implement modifications and customization for different user preferences.
-
26:00 🚗 Tesla employee comments on the crossover expertise in neural nets between FSD and Optimus projects.
------------------------------------- 0:27:9 2024-01-09T19:46:56Z
Review
This content centers on Tesla's Full Self Driving (FSD) Version 12 (V12), offering an in-depth look at its advancements, testing, and employee feedback. Here's a comprehensive review of the content:
Overview of Tesla FSD V12
- Improvements: Tesla's Full Self Driving V12 showcases significant enhancements in human-like driving behavior, proficiency in navigating roundabouts, and overall stability.
- Testing: Tesla employees are actively testing FSD V12, with the rollout occurring in phases, starting from executives to team leads, and then to the Early Access program for employee-owned vehicles.
- Employee Feedback: The majority of Tesla employees have noted improvements in aspects like lane keeping and roundabouts, with a general consensus on the enhanced performance of FSD V12.
Key Insights
- Investor Expectations: There's a high anticipation among Tesla investors for the wider release of FSD V12, with expectations of a substantial improvement in technology.
- Neural Net Advancements: The FSD V12 is completely based on neural networks, learning from video data to enhance human-like driving behaviors.
- Self-Learning Capabilities: The system's ability to learn and improve over time is emphasized, indicating a continuous evolution in its performance.
- Confidentiality of Beta Tester Feedback: The feedback from FSD beta testers, considered crucial in refining the system, is highly valued yet kept anonymous.
Clips and Time-Stamps
The content includes specific clips with time-stamps, likely from a video source, providing detailed insights into various aspects of FSD V12. These include:
- Feedback from Tesla Employees: Discussing improvements in FSD V12 and the need for testing in diverse environments.
- Neural Net Implementation: Focusing on how neural networks contribute to making the driving experience more humanlike and safer.
- Weather Condition Testing: Highlighting the need to test FSD V12 in various weather conditions.
- Adaptability and Safety: Stressing the importance of ongoing learning to adapt to different road conditions and the aim of achieving impeccable system safety for autonomous driving.
- Release Timeline: Discussing the potential release timeframe for FSD V12, with possible delays due to other Tesla priorities like the Cybertruck.
- Tester Impressions: Sharing experiences and feedback from both Tesla employees and experienced FSD beta testers.
- Expansion and Customization: Detailing plans to expand FSD to more customers and using neural networks for easier system adaptation and customization.
- Crossover Expertise: Mentioning the shared neural net expertise between FSD and other Tesla projects like Optimus.
Conclusion
The report on Tesla's Full Self Driving V12 paints a picture of a cutting-edge autonomous driving system that is undergoing rigorous testing and refinement.
The emphasis on neural network technology, combined with real-world feedback and continuous learning, suggests that Tesla is on the verge of a significant breakthrough in self-driving technology.
The detailed feedback from employees and testers underscores the system's potential and areas for improvement, hinting at a future where Tesla's FSD technology could play a pivotal role in advancing autonomous driving.
--------------------------------------------------------------------------------------------
Top comments
People are expressing concerns about the functionality and capabilities of Tesla's Full Self Driving V12, with some feeling that it is not yet ready for widespread use
-
If I am not wrong I though v11 supposed to be all NN.
-
I hope this video helped you get brighter about Tesla! My website: https://www.herbertong.com See the MOST comprehensive resource for the $TSLA Investor Get FREE TESLA Milestone Tables Join this…
-
When I use FSD beta my biggest override is pressing the accelerator at stop signs. This is a regulator corrupted issue and needs to be removed, it causes other drivers to not know what to do. Lane…
-
SYMBIOSIS! I haven't heard anybody address this in a meaning full way. If tomorrow the rules of the road were all changed to meet the capability's of the current FSD then PROBLEM SOLVED. Build the…
-
As a truck driver, the best place to learn to drive is Atlanta, Texas Central and North Illinois, Wisconsin and maybe New York City and a little bit of Jersey if you can drive there, you can drive…
-
Bottom line. I work hard to train FSD and paid $10,000 for privilege because I am a Tesla investor and want it to work to increase my investment. If I purchased FSD to help me drive or just be user I…
-
It seems that FSD is not solved with v12 😐
-
WS, MM do not believe FSD will be robotaxi capable this decade. Based on never ending beta's, massive time misses with no ability to improve predictions who can blame them? The problem is even if FSD…
-
Great and engaging conversation. Surely the driver ought to be able to adjust the rain sensor threshold so his car knows what behaviour and level of service to aim for
