Mercedes becomes the first automaker to sell autonomous cars in the U.S.

Original Article ↗ Hacker News Discussion ↗

Mercedes has begun selling SAE Level 3 “conditionally autonomous” cars to consumers in California and Nevada, making it the first automaker in the U.S. to assume legal liability when its system is driving. Commenters debate how meaningful this is given the system’s narrow operating conditions—daytime, mapped freeways, heavy traffic under 40 mph—contrasting it with Tesla’s more broadly capable but driver-supervised Level 2 system and Waymo’s fully driverless robotaxis. Many see Mercedes’ willingness to take responsibility for crashes within its defined domain as the real breakthrough, with the potential to shape regulation, consumer expectations, and the path to scaling autonomous driving.

Scope of Mercedes Drive Pilot

  • System is SAE Level 3, but with a very narrow operational design domain (ODD): daytime, heavy traffic, specific pre‑approved California and Nevada freeways, under ~40 mph.
  • Only ~65 vehicles were reported available in California; seen as “testing the waters” rather than a full rollout.
  • Some see the limited ODD as nearly “parking-lot narrow,” making headlines about “autonomous cars” feel misleading or PR‑driven.

Liability and Meaning of Level 3

  • Key distinction raised repeatedly: Level 3 shifts legal responsibility to the manufacturer within the ODD; driver need not continuously supervise.
  • Mercedes is said to take liability, with caveats about proper use and maintenance; some commenters doubt how watertight this commitment is in practice.
  • Takeover requirement is around 10–12 seconds after warning, which some consider sufficient, others see as a gray and risky area.

Comparisons to Tesla, Waymo, and Others

  • Many contrast Mercedes L3 to Tesla’s “FSD”/Autopilot (Level 2):
    • Tesla operates more broadly (city streets, highways, higher speeds) but requires constant attention and retains driver liability.
    • Several users report Tesla FSD as stressful, requiring frequent interventions, and unsafe in edge cases; others claim it’s far ahead technically but can’t assume liability.
  • Waymo and Cruise are cited as Level 4 robotaxis:
    • Waymo is praised by regular users in SF/LA as highly reliable and effectively replacing Uber.
    • Others argue Waymo has “failed to scale” relative to 2017 expectations; countered by claims of steady multi‑city expansion.
    • High sensor cost and ongoing unprofitability are noted; margins per ride are unclear.

Usefulness and User Experience

  • Some see Drive Pilot as highly valuable precisely where driving is worst: low‑speed, stop‑and‑go traffic on congested freeways.
  • Others dismiss it as “self‑insured cruise control” with marginal benefit compared to existing advanced driver‑assist systems.

Broader AV Trajectory and Strategy

  • Debate over whether incremental L3 with tight geofencing (Mercedes) or ambitious, map‑light generalization (Tesla) / high‑sensor robotaxis (Waymo) is the better long‑term path.
  • Several note that AV timelines from the late 2010s were overly optimistic; nonetheless, many think steady progress continues, with 2030 often implied as a more realistic horizon.