I especially enjoyed hearing about JT‘s past experience at Tesla and his early interest in electric vehicles. We ourselves purchased a Nissan Leaf as soon as our reservation came up, and we later purchased a 2012 Model S (used, from Tesla).
However, I am not on board with some of the negative commentary on Autopilot. While we still have our old Model S (at over 200K miles), we appreciate using our Model Y on longer drives precisely because of Autopilot/FSD. The Model S is Tesla‘s best looking car, in my opinion, and I still love the way it drives, but I have found Autopilot to be very convenient and helpful for safety on long highway drives.
Also, as one with a software background, I have a hard time understanding why anyone thinks eight cameras are not enough. That assertion was made during the podcast, but no real support for such a position was provided. It seems to me that the video from eight cameras is more than sufficient to monitor the surroundings and road in all directions. As an FSD beta tester, I am more inclined to believe that the biggest challenge is not necessarily perception, but rather intelligent decision making. While using FSD, I have found myself in countless situations where the car’s perception seems to be fine, but the driving logic is flawed. For a simple example, when turning right onto roads without lane markings, the car often swings too wide, which can spook other drivers. It also tends to drive too far to the left on roads without lane markings, and this is reflected in the FSD display, indicating that the problem is one of “driving intelligence“ rather than perception.