* The given description of Tesla's "Shadow Mode" has been bouncing around for a while, but it's not accurate (see https://twitter.com/greentheonly/status/109632

* The given description of Tesla's "Shadow Mode" has been bouncing around for a while, but it's not accurate (see https://twitter.com/greentheonly/status/1096325954375430144 https://twitter.com/greentheonly/status/1096325954375430144 for a description of what Tesla actually does). * The description of LIDAR having "much more detail than other sensors" is confusing. LIDAR has more detail than radar, but LIDAR is much less detailed than cameras. * The description of the 2016 accident is not correct. MobilEye provided the EyeQ3 chip on the car, not just the software. "Since then, Tesla added radars for these situations" is flat out wrong; Tesla has always had radar as part of AP, and radar didn't (and won't) prevent that sort of accident (https://twitter.com/artem_zin/status/748623772056428544 https://twitter.com/artem_zin/status/748623772056428544) * 'the newest Tesla Autopilot software 9.0 has the largest vision neural network ever trained' wut? * 'The DARPA Grand Challenge began in 2004'...'the winner, Waymo' wut? * 'systems based on if-then statements' ? * 'Comma.ai continues to develop its software, and drivers can buy and install it in their cars' the software is free, it's on Github. Furthermore, the thesis in the HN title is not supported at all in the post as far as I can tell.