Conversation

Scale faster? Why do you think this matters? What's the first mover advantage? It's not a marketplace and there's no network effect. It behaves like scooters, not like Uber or Facebook. Many investors are going to be very disappointed, esp when they learn the price of openpilot

Quote Tweet

Rogs

@ESRogs

Apr 11, 2022

Replying to @Theophite and @ModeledBehavior

Yes. And humans do it, so we know it's possible. It's possible that Tesla (and comma.ai) won't be able to make the models good/fast enough in time to scale faster than all the LIDAR approaches. Time will tell.

12:53 AM · Apr 12, 2022

Replying to

Yeah, if investors had an option to buy Comma, maybe they wouldn’t have to be disappointed.

Replying to

Engineers are the future, not investors. History is a progression of the value moving from those who have capital to those who can do things.

Show replies

Replying to

Now Elon has basically conceded his current Tesla fleet will never be above L2 (why would he need the new dedicated futuristic robotaxi?), what is your projection when his new dedicated futuristic robotaxi will achieve L4 like Waymo, Cruise and others?

Replying to

L5 exists today and works well. Hire a driver.

Show replies

Replying to

Is it cuz it’s mostly a hardware thing and not a software thing, so it won’t scale as fast as social media buy it can scale as fast as smartphones?

@ykssaspassky

Apr 12, 2022

Replying to

@comma_ai

It’s beginning to be clear that a solution of FSD is actually an AI project rather than a driving project.

Randy O'NeillCylinder

Replying to

Charles R Qi

changing from RF to a DL standard with pointnets for Lidar. C'mon these kind of mindsets are just super cool. And with e.g. spacecraft docking maneuvers without the sun, there seems to be some real benefit.