High Self Driving System Analyst Amazed by Tesla 12.5.2.1 Beginning to Go Past Human Driving

High Self Driving System Analyst Amazed by Tesla 12.5.2.1 Beginning to Go Past Human Driving
High Self Driving System Analyst Amazed by Tesla 12.5.2.1 Beginning to Go Past Human Driving


After months with FSD 12, I’ve gone previous the “wow, that is magic” part. The analyst in me scrutinizes each disengagement and edge case. However at this time, I simply had a type of “tears in my eyes” moments.

On a mini street journey, FSD 12.5.2.1 flawlessly dealt with the winding roads which have at all times been my worst nightmare. Actually no human I do know might’ve finished higher. Completely centered, easy turns, excellent pace. For 2 hours, the motive force was fully relaxed, completed breakfast, and even probably the most carsick passenger performed on their cellphone—zero disengagements, completely easy. It handed the “espresso cup check”.

Reaching AGI in self-driving is among the hardest challenges on the market, however with 3 trillion miles pushed yearly and 12 hours per week behind the wheel, it’s poised to be one among humanity’s best unlocks. We’re speaking trillion-dollar disruptions that nobody’s totally greedy but—reshaping cities, actual property, visitors, and automotive possession!

Each step ahead will get us nearer to that future all of us deserve.

And at this time, I felt it.

Earlier Testing of China and US Self Driving

In July, 2024, Freda ust completed a one-week journey to China. She survived all the main (~20) L2 self-driving and robotaxi automobiles in each the US and China. Some ideas & observations:

Freda Duan, Altimeter Capital, examined main manufacturers like $Huawei, $Li, $NIO, $Xpeng, and $Xiaomi. Total, they exceeded my expectations. The rides weren’t overly cautious and dealt with complicated conditions (sure, street circumstances in China are very difficult!) fairly effectively.

Nothing compares to $Tsla’s method. I see imitation studying/end-to-end as the one efficient method for self-driving. Whereas Chinese language friends carry out effectively on important roads, they battle on frontage roads as a consequence of reliance on high-precision maps and rule-based strategies (e.g. vehicles stopped in the course of the street the place there was no clear white lining).

Chinese language EVs’ self-driving capabilities are far forward of these from US and EU manufacturers.

I doubt any Chinese language gamers can revenue from L2 self-driving, not as a result of it’s not helpful, however as a result of it’s laborious to distinguish, and value wars dominate the market in China.

Chinese language customers and regulators appear way more receptive to self-driving. Even with a 5/10 self-driving functionality, vehicles are virtually *hands-free(!)*

Insurance coverage-wise, for L3+ vehicles, OEMs bear accountability for incidents, so OEMs keep away from labeling vehicles as L3+.



Leave a Reply

Your email address will not be published. Required fields are marked *