SF public transit sits at half a million per day ridership so I'm looking forward to being crammed into a car with about 333 other people
Instead of 1000 buses what about 15,000 smaller buses without routes or schedules that would be much more efficient
Wild how at this point we are arguing for busses instead of bazinga rental scams; when less than a century ago busses were the bazinga scam getting rid of trams, railcars, and trolleys.
its a fuckin real life strawman
"See how bad and expensive transit is?? (because we bulldozed the good stuff and purposely made it bad) We should do [Bazinga bullshit] instead"
conservative 12 year average life of the waymo
...what? Taxis do like 50k+ miles a year of hard city driving, they get replaced every 3 to 6 years. So even his "conservative" math is fucked.
His sole piece of evidence for the longevity of an urban taxi cab is the one driven by Earnest Borgnine in Escape from New York
Tom Nicholas did a good video on how this company buys up science youtubers to pump their company to others as a thinly veiled "cool science video on AI cars" and it's pretty damning. Veritasium crying now
https://youtu.be/CM0aohBfUTc
The one veritasium video I watched was about the Incompleteness Theorem. Since it was just about a math topic I don't think there was anything to shill, and I thought it was well done (not a mathematician, cannot assess it in that way). But then I see the other videos recommended now and again and they typically look like bait 'n bro.
Theorem: Any lay explanation of Gödel's incompleteness theorems is either incomplete or inconsistent.
Proof: :made-it-the-fuck-up:
I'm only sour about these because it's one of those things like string theory or "0.999...=1" that attracts almost exclusively cranks who want to use whatever details they internalized from the NATOpedia page and use it to explain why orgones are real or whatever.
I know I'd need to do a deep dive to actually understand the theorem, so I've trained myself to actively reject any information about it because I assume it would be wrong.
Are you familiar with the Metamath project? https://us.metamath.org/mm.html
They don't have a complete proof of Godel's incompleteness theorems yet but I feel like I must plug them anyway lol
It's an ongoing attempt to express and formalize all of mathematics via a massive collection of theorems defined as rules to rewrite the basest axioms of formal mathematics into the theorems to be proved (although in practice you usually start with your theorem and work backwards in Metamath). For any theorem in the database you want to get an understanding of, you can look at the rewriting rules which are expressed as a series of steps to understand why a theorem is true. Or if something is hard to believe you can at least look at the computer-verified proof and safely accept a theorem as true by the rules of the system :3
Has been rly useful to me as someone interested in learning about abstract math but not having a place to start
For example here are the proofs for 0.999 = 1 and 2 + 2 = 4
https://us.metamath.org/mpeuni/0.999....html
https://us.metamath.org/mpeuni/2p2e4.html
I'd heard he was good but my first taste was that video and it's like 💀 not even subtle shilling.
If you're a very math oriented person I think you probably wouldn't like anything he makes for being too surface level. If you're interested in math but haven't done math beyond maybe an undergrad level, his math history videos are both gorgeously animated and pretty compelling in narrative so they help carry across the concepts. All the other videos tend to be slop.
Yeah it would probably be lost on me but it's absolutely something my bf would love so I'm recommending it to him
His channel has settled into a rhythm of doing 1 really good, gorgeous video on math or physics, then a horrible slop video that's often just a commercial for some company. If you can pick up on the vibe of which one you're watching at any time, you need only waste 30 or so seconds before you skip the ones that aren't worthwhile.
I found a YouTube link in your comment. Here are links to the same video on alternative frontends that protect your privacy:
Against:
-
Ridiculously more energy inefficient
-
Incredibly space inefficient
-
Incredibly throughput inefficient
-
Will be much more expensive than using public transport.
-
Will cause a lot more traffic
-
Will cause a lot more accidents
-
There's no way those operating costs are correct
-
I worked for them for several years and $8,000 in operating costs per vehicle is utterly ridiculous
Electric cars do require much less maintenance, but 8K a year would barely pay for someone to keep the windshield wiper fluid topped off much less the kind of issues that would pop up in a fleet of 15k vehicles.
I wonder what would happen to traffic with 15,000 extra vehicles within the city limits. Something tells me it won't relieve congestion or make the streets any safer.
But it would make a couple of vulture capitalists a lot of money.
I mean, even on top of maintenance the entire system is basically just on rails, it has to be mapped out manually, every car has a 'guy in the chair' that can control the car remotely if necessary, they also got guys in unmarked vans that follow the driverless cars in case of emergency, and on top of that every car is constantly collecting data and that storage is insane, all of that falls under operational costs completely independent of maintenance issues
yeah it sounds like that dril tweet about needing help with his budget but for coming up with operating cost assumptions for a business plan
please help why won't any venture capitalists fund me?Wouldn't even cover the human fluids cleanup costs of 15,000 autonomous taxis
I don't think anyone here needs convincing about why self driving cars are a terrible idea, but Not Just Bikes posted a video about it some days ago laying out how dire it really is if we allow self-driving cars to go on unquestioned
yeahhhh... mixed feelings on that vid. it was highly speculative but still more or less compelling. I think the future will be both dumber and less interesting than what he proposes, but the undeniable fact is they can pretty much only make things worse
I agree, it is really hyperbolic, and it's likely not to go that way. But it lays out the limitations and dangers of giving the streets to private companies and having transit be held hostage.
Baidu is the largest robotaxi operation in the world right now and China doesn't tend to let companies burn money on nonsense for too long, so I think they at least have a spinoff merit.
Chinese people are not immune to techbro brainworms though, just look how popular Melon and Tesla are in the PRC.
I believe the cpc has members on the board of Baidu, so this isn't about the people, this is about the state.
Wtf is a Waymo. Sounds like a kids toy.
Tickle Me Waymo lmao
It's a self driving taxi startup. It doesn't drive itself reliablh. It relies on call center workers in the global south to take control of the cars when they run into trouble. They already have killed several people and are programmed to only avoid specific objects they identify, if they identify you as 'other' they'll just run you over (yes this has already happened)
Anyway, paying a bus driver and building rail infrastructure is completely unthinkable and inefficient dontcha know?
They already have killed several people
I have been unable to corroborate this with any sources. To me it seems like most of their collisions are the result of people not paying attention and rear ending them at stoplights. I also heavily doubt that they only avoid known objects because of how easy it is to imprison them by putting random things in their vicinity and by their comically incremental right turns.
Wait you're completely right, Waymo hasn't had fatal crashes. I mostly was going off this video he mentions how they don't avoid things they don't recognize (maybe the algorithm has changed)
Yeah, that makes sense, especially with the object permanence issues for cruise. I've worked on similar systems with vision data and figuring out how an object leaves frame is significantly easier said than done. My big qualm with that video is the degree it minimized the danger of human drivers. The advantage of autonomous vehicles isn't traffic reduction or price or any of the topics he mentioned; rather it's that machines don't get drunk or high or tired or angry. Autonomous vehicles aren't a permanent solution, however, as a concept they serve as harm reduction by being significantly safer than humans when comparing mileage.
I found a YouTube link in your comment. Here are links to the same video on alternative frontends that protect your privacy:
Weird that "it is a functioning product" isn't listed in the assumptions
Good points here on operating costs. Want to add, there is no way one of those is 120k. The shitty jaguar truck it's based on is at least 60k alone delivered and then it needs to get totally disassembled to install the arrays and sensors to make it
run over pedestrians on purposedrive.Because that’s what California needs! More traffic! Does it come with a free AI generated printed image of a crying train too?
Can other STEM disciplines get as smug as CSers are? I’d love to have a civil engineer and an urban planner laugh at these dipshits for claiming we should abolish public transport for their technoslop.
Can other STEM disciplines get as smug as CSers are? I’d love to have a civil engineer and an urban planner laugh at these dipshits for claiming we should abolish public transport for their technoslop.
Oh they can, in my experience mechanical and aeronautical engineers do it a lot. Not to mention chemical engineers who have a very material reason to be kissing big oil and big processed food's ass.
Engineering school is where humility goes to die, the worst part about chemical engineering school was dealing with the overinflated egos of everyone, from professors to freshpeople.
Shit sucks, engineering sounds like you'll either come across the best people you'll ever meet, or ghouls who knew they were smart enough for something beyond MBAs.
Yeah, I don't have any friends from engineering school anymore, but it's also because I took such a radically different path after graduating that we just kind of fell off. I know some engineers who seem to be cool, but no one I know for sure.
Actually 15,000 is more than enough to cover most of the average number of people taking public transit per hour
I've stopped reading after the first sentence and am just letting the idea of 15,000 waymos in one place marinate in my mind
Clown car
