Vision-only autonomy is incredible. But it isn’t flawless. If AI can learn from text, images, audio, and video — why is adding LiDAR treated like heresy? At Rivian Autonomy Day, something interesting happened. Instead of declaring a single “correct” path to self-driving, Rivian doubled down on sensor fusion: cameras, radar, and LiDAR working together. Not because vision can’t work — but because autonomy is too important to be solved by ideology. This debate feels eerily familiar. In the War of Currents, Thomas Edison defended DC while Nikola Tesla pushed AC. Progress didn’t come from silencing alternatives — it came from letting better ideas compete. Today, the autonomy space risks repeating that mistake by shutting down parallel approaches instead of encouraging them. This video breaks down: • Why vision-only autonomy works — and where it fails • What LiDAR actually adds (and what it doesn’t) • Why redundancy and multiple sensing paths matter for safety • How Rivian and Tesla represent two fundamentally different philosophies • And why the future of autonomy needs competition, not dogma This isn’t about picking sides. It’s about asking the harder question: Should autonomy be solved by belief — or by evidence? Let multiple approaches compete. Let reality decide. May the best system win.
from Tesla In The Gong https://www.youtube.com/watch?v=G89aIjSVCiI
No comments:
Post a Comment