FSD比人驾驶更安全?专家质疑特斯拉报告:数据存在缺陷有误导性

Core Viewpoint - Tesla has launched a new section on its website to report safety statistics for its Autopilot and Full Self-Driving (FSD) systems, but safety experts criticize the data as flawed and misleading [1][2]. Group 1: Safety Data and Performance - Tesla reports that FSD users have driven 6.47 billion miles, with a serious collision occurring once every 5.1 million miles and a minor collision every 1.5 million miles, significantly better than the average American driver [2]. - The previous quarterly safety reports were criticized for focusing primarily on the Autopilot system used on highways, neglecting the more common accidents that occur on city streets and non-divided roads [2]. Group 2: Expert Criticism - Experts like Philip Koopman from Carnegie Mellon University acknowledge the new report's distinction between highway and non-highway mileage as a positive step, but argue that the data undermines Tesla's claim that using FSD is safer than not using it [3]. - The report lacks information on injuries or fatalities related to Autopilot or FSD incidents, with Tesla claiming that such data relies on voluntary reporting from drivers, which raises skepticism among experts [3][4]. - Experts express concerns that Tesla's safety documents are filled with marketing hype rather than serious safety analysis, contrasting Tesla's approach with that of Waymo, which provides peer-reviewed studies to support its safety claims [4]. Group 3: Trust and Verification Issues - Engineers like Noah Goodall find it challenging to publish research on Tesla's data due to doubts about its authenticity, citing Tesla's history of misleading the public regarding safety data [5].

FSD比人驾驶更安全?专家质疑特斯拉报告:数据存在缺陷有误导性 - Reportify