Tesla FSD Timeline
Repeated promises that Tesla’s “Full Self Driving” (FSD) is just a year or two away are contrasted with its current limitations, recalls, and failures on basic tasks like speed-limit recognition and phantom braking. Commenters debate whether this gap between marketing and reality is overoptimistic vaporware or outright fraud, especially given that customers have paid thousands for FSD years ago without receiving hands‑off autonomy or robotaxi capabilities. The thread also touches on technical choices such as “pure vision” versus lidar and the broader stakes for safety, regulation, and Tesla’s valuation if true Level 4–5 autonomy remains distant.
Perpetual Timelines & Credibility
- Commenters tally years of “FSD is ~2 years away” claims and see a pattern of missed, repeatedly shifted timelines (robotaxis by 2019, Level 5 by 2021, etc.).
- Many argue this isn’t just “optimistic” but systematically misleading, especially given how central FSD has been to Tesla’s valuation and marketing.
Fraud vs Over‑Optimism
- One camp calls it outright fraud: money taken for a non-existent feature, with specific timelines and revenue promises (e.g., cars earning as robotaxis).
- Others frame it as entrepreneurial overconfidence and “vaporware,” noting Tesla has invested heavily in FSD and does show incremental progress.
- Debate over legal/ethical lines: when does “bold prediction” become false advertising or market manipulation?
User Experience: Enthusiasts vs Burned Buyers
- Some owners say FSD/Autopilot is “pretty damned good,” use it daily, and view it as futuristic and impressive despite quirks.
- Many others report serious dissatisfaction: FSD packages bought years ago that still “do practically nothing,” especially outside North America (e.g., can’t read LED or variable speed signs in Finland, Norway).
- Basic assist features (wipers, high beams, cruise control without phantom braking, parking/summon) often described as unreliable or regressed; some want refunds or mention consumer-protection actions.
Technical Approach: Vision‑Only vs Sensor Fusion
- Strong criticism of Tesla’s removal of radar/ultrasonics and rejection of lidar; several argue safe autonomy “needs” lidar + vision and classic sensor fusion.
- Others defend pure vision as ultimately more scalable and “human-like,” claiming added sensors increase complexity with limited payoff.
- Several note that sensor fusion (e.g., Kalman filters) is standard robotics, not a deep unsolved problem; the core challenge is robust perception and edge cases.
Safety, Data, and Benchmarks
- Concerns about crashes into static objects (including the parked airplane video), phantom braking, and Tesla keeping detailed performance data private.
- Suggestions to track interventions over fixed routes, but worries about Goodhart’s law and Tesla optimizing for benchmarks rather than real-world safety.
Comparisons & Broader Value of FSD
- Waymo in San Francisco is cited as an operational, geofenced “FSD-like” service; Tesla is praised for generality but criticized for weaker safety envelope.
- Some argue FSD for personal cars is a “solution in search of a problem”; others see huge value for elderly, disabled, or non-drivers and freight/logistics.
- Repeated analogies to fusion power, Theranos, and cults of personality highlight concerns about hype outrunning reality.