Comments by "Nicolae Crefelean" (@kneekoo) on "Self Driving Tesla Causes 8 Car Pileup" video.

  1. Man, I'm all for crapping on bad marketing, hype, and misplaced optimism. But words matter, and it doesn't help putting words in his mouth. He's very optimistic and all that, but so far I haven't seen him promising a specific date for level 5 FSD - and I watch critic reviews and interviews because I'm interested in how this effort goes. Planning/targeting something doesn't guarantee success or making it in time. It can easily mean that despite all effort, the plan fails. Tesla managed to come up with some stuff on time from their master plans, and other stuff was either delayed or *very* delayed. FSD is the latter case. To be fair, no one has it, so you can't really estimate based on anyone else's timeline of solving it. That's why despite me rolling my eyes every time I hear them saying they're getting closer, I keep reminding myself that if this were easy enough, others would've done it already. Now clearly they're both far ahead on their FSD work, and far from level 5 autonomy. FSD still feels to me like something likely in medium term, at best, because there's just so much that can go wrong. And that's why I don't like that anyone can get the FSD beta now. Even before they opened it, as shown, some people are plain irresponsible. I'm not sure what to make of this public road testing. If we look at the numbers, considering Tesla cars have less crashes per million miles, compared to the national US average (according to NHTSA), we can easily argue that shit happens for everyone, including (but less) with Tesla cars. This is why I don't know what to make of this - the numbers are in their favor, even with this kind of crap. And it feels like people expect Tesla to either be perfect or stop testing FSD publicly. Sandbox FSD can't become real-world FSD faster, because there are plenty of edge cases that you might never learn in a simulated environment. So you pretty much have to make public tests. But I would definitely keep it a closed beta. And frankly, I find it mind-boggling that you have to pay 15k to be a beta tester. But I guess you can't tell people what to do with their money, so there's that.
    1