Analysis

Waymo Founder Throws Shade at Tesla's 'Vision-Only' FSD: Calling It a 'Bad Case of Myopia'

Post on XShare on Facebook
Waymo Founder Throws Shade at Tesla's 'Vision-Only' FSD: Calling It a 'Bad Case of Myopia'

TL;DR: Waymo founder John Krafcik just dropped some truth bombs on Tesla's 'vision-only' Full Self-Driving, calling it out for having 'a bad case of myopia.' He says relying solely on cameras without LiDAR and radar is like handcuffing the AI, especially when others are building 'superhuman' systems. Ouch! He's got receipts, too.

Meta: Waymo founder John Krafcik sharply criticizes Tesla's 'vision-only' Full Self-Driving approach, calling it a 'bad case of myopia' and arguing that the lack of LiDAR and radar inherently limits its safety and reliability.

Alright, alright, gather 'round, folks, 'cause John Krafcik, the man who helped build Waymo, just went in on Tesla's Full Self-Driving, and he didn't hold back! You know the debate: Elon Musk says cars should drive like humans, with just cameras and a brain. Krafcik and most of the industry are screaming, "Hold up!" saying you need LiDAR and radar for true safety and redundancy. And now, Krafcik's gone on record, calling Tesla's FSD a "bad case of myopia." That's a fancy way of saying it can't see straight, even with glasses!

Krafcik laid it out plain: human eyes are way better than a car with a few 5-megapixel cameras spread out. He says when you disperse those megapixels, the effective vision is more like 20/60 or 20/70. Your car wouldn't even pass a DMV eye test! He's arguing that by ditching radar and ultrasonic sensors years ago, and refusing to add LiDAR, Tesla has basically tied one hand behind its AI's back. Meanwhile, folks like Waymo are out there with LiDAR and radar, getting different types of data, building what he calls "truly superhuman" systems. That's a punch to the gut for the 'vision-only' believers.

A Track Record of Truth

Now, you might say, "Eddie, why should I listen to him?" Well, this ain't Krafcik's first rodeo. Remember back in early 2025 when he predicted Tesla would "fake" its Robotaxi launch? He said they'd just mimic a working service. And what happened? When Tesla's Robotaxi pilot rolled out in Austin, it was confirmed they were leaning heavily on remote monitors and safety drivers. He was spot on! Tesla keeps missing deadlines for removing safety monitors, and Krafcik's saying it's because the hardware itself is the problem. No amount of software magic can fix a camera blinded by the sun or caked in mud.

This is a real head-scratcher for Tesla owners, especially those who shelled out big bucks for FSD. If Krafcik's right – and his track record suggests he might be – what happens to all those vision-only Teslas out there, hoping for Level 4 autonomy? It raises serious questions about the fundamental approach. It's physics versus software, and right now, Krafcik says physics is winning.

What's Next

This debate ain't going away. As autonomous tech evolves, the industry will keep watching to see if Tesla's vision-only bet pays off or if a multi-sensor approach becomes the undisputed champ for true Level 4 autonomy. Tesla's gonna have to pull some serious rabbits out of hats if they want to prove Krafcik wrong.

For now, Waymo's got real robotaxis on the road, and Tesla's still got supervisors in the driver's seat. Scoreboard's leaning Krafcik's way, folks. Truth hurts, don't it?

Comments

Join the discussion below.

Loading comments…
Eddie W

Eddie W

Author

Need an OG image?

Share this story to automatically generate an image via /api/og.

We’d love your thoughts

Help steer EV-Insider by sharing what you want to see next.

Few quick questions, no fluff. Tell us which stories spark your interest and how we can make this daily brief more useful for you.