FSD v14 marks a leap in Tesla's autonomy, delivering end-to-end driving with human-like decision-making that outpaces prior versions in perception and response.

Key Takeaways

• Enhanced beginning and end-of-drive capabilities allow seamless backing out from tight spaces and parking in varied spots.

• Rapid reaction times to environmental changes, like ambulances or pedestrians, improve safety but introduce brief brake taps for caution.

• Emergent behaviors emerge, such as navigating drive-thrus by detecting transactions via cameras, without specific training.

• Hesitancy in lane changes and speed profile adjustments signal a shift toward more conservative, context-aware driving.

• Path to SAE Level 3 autonomy appears close, with new alerts for increased attention paving the way for relaxed supervision on highways.

This update transforms Tesla vehicles into more sentient systems, excelling in unprotected turns, obstacle avoidance, and low-speed maneuvers. While smoothness issues like abrupt braking persist, the core advancements in neural network processing enable quicker adaptations to dynamic scenarios, from railroad crossings to construction zones. Hardware 4 models benefit from front cameras for precision, yet hardware 3 compatibility remains a focus for broader rollout. Overall, v14 edges closer to robotaxi viability, prioritizing safety over aggressive actions, with future iterations likely refining speed limits and parking logic for polished performance.

Autonomous ride-hailing is exploding, with Waymo's established scale clashing against Tesla's aggressive beta rollout—revealing key edges in fleet growth, ride quality, and market disruption that could redefine urban mobility for tech-savvy users.

Key Takeaways

• Waymo operates over 1,500 vehicles across four cities, logging 250,000 paid rides weekly, while Tesla's Austin fleet sits at 11 Model Ys with safety monitors, targeting 30+ vehicles soon.

• Tesla rides feel superior in smoothness and infotainment sync, outperforming Waymo's jerky stops and unprotected left turns observed in tests.

• Waymo commands 25% market share in San Francisco, overtaking Lyft despite 30-40% higher prices than Uber, driven by privacy and no-tip appeal.

• Tesla plans geofence expansions north of Austin's river, tackling pedestrian-heavy zones like Sixth Street, while prioritizing fleet growth before unsupervised ops.

• Long-term, Tesla's visionless approach and billions of FSD miles enable faster scaling than Waymo's mapped, sensor-heavy method, potentially undercutting prices at 5-10% below Uber.

The discussion kicks off with Waymo's impressive footprint: 1,500 vehicles churning out 250,000 rides weekly across Phoenix, San Francisco, Los Angeles, and Austin, backed by Alphabet's deep pockets and partnerships like Uber for fleet ops in new markets. This scale took a decade of incremental mapping and sensor integration, yielding a service that's reliable but not flawless—rides often feel abrupt with hesitant braking and occasional aggressive merges, as noted in back-to-back tests against human-driven options. Despite these quirks, Waymo's captured 25% of San Francisco's ride-hailing gross bookings by April 2025, eroding Uber's dominance from 63% to under 55% and matching Lyft's share, even at premiums of $5-6 over Uber's $15.58 average fare. Riders cite the no-driver privacy and zero tipping as offsets, with 70% in surveys preferring the experience for comfort over cost.