Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register
This site may earn commission on affiliate links.
Right? I mean they don’t even need FSD Beta for this. They control the lanes, the lineage, the markings, the signaling - everything. Just train up a vision stack on every inch of the tunnel system (could do this with simulation) and write straightforward procedural code for everything else. Don’t even have to get regulatory approval.
Yes, it doesn't need to be all that fancy. They could paint a line down the middle of the tunnel and have a 2¢ robot program drive it. Like every factory robot everywhere can already do. They really should get their vision FSD properly written to do it though, it's amazing that they don't care enough to fix the problem.
 
Yes, it doesn't need to be all that fancy. They could paint a line down the middle of the tunnel and have a 2¢ robot program drive it. Like every factory robot everywhere can already do. They really should get their vision FSD properly written to do it though, it's amazing that they don't care enough to fix the problem.
Then one of the cars breaks down, a passenger wearing some crazy outfit that the perception neural net doesn't recognize as human gets out, walks down the track, and gets mowed down by the next car. The system has to be extremely reliable.
 
Interesting, I was assuming that the upcoming robotaxi model would be a robovan. Maybe the robotaxi model will be small?
I think the emphasis should be on “highly configurable” in this tweet. In any case they need to finalize the design, cannot believe they are still fiddling with these details; there is not much time before the first passengers!

Looking forward to the “one more thing” reveal of the finalized production-ready Robovan/Robotaxi at the end of AI Day!
 
Then one of the cars breaks down, a passenger wearing some crazy outfit that the perception neural net doesn't recognize as human gets out, walks down the track, and gets mowed down by the next car. The system has to be extremely reliable.
Yes, automated driving in the Boring tunnels should be extremely reliable. They have the resources, so they can make it happen. The upside to getting autonomy is they can eliminate 70 drivers in the Vegas Convention Center loop, and hundreds more in the eventual Vegas Loop. Plus eliminate this public embarrassment and fulfill their design goal (and implied promise) of autonomy in the tunnels.

It seems an odd project to leave unfinished.
 
  • Like
Reactions: 2101Guy
Why/what was The benefit to him of waiving due diligence clause before signing the contract?

Was this a waste of (at least) a billion dollars for absolutely nothing/poor decisionMaking on his part? Or am I missing something?
I consider it a >$40B savings since in all likelihood in 10 to 20 years Tweeter will be as irrelevant and worthless as MySpace.
 
GREAT news.


The title made me laugh. Someone is having too much fun over at Tech Crunch.

 
That is not bold. Pretty bold of Elon though. I guess it doesn’t matter if he is right or not. Also “handle” could mean working 10% of the time (as compared to prior 0%).
That's ridiculous. 95% is still horrible for a geofenced test that requires many 9's to achieve human parity.

I still want Chuck to measure how often humans fail on that left turn.