Should Tesla start it’s own taxi network driven by owners?
Tesla FSD Beta v9 is out, using Vision only neural nets - intended to perform all the moonshot Lvl 5 tasks. It seems logical for them to have human drivers do Uber stuff, learn from it - learn the last mile/parking lots better- train DL modes to get better at those nuanced scenarios; and then launch a Robo taxi v1 based on that. Some but not most owners would probably enroll in a taxi network to drive themselves. (Still not a poor people car) Nevertheless, some will enroll in the option to convert your Tesla to an Uber. Their Beta users are a anyway ~1000 using FSD beta. So they’re okay with testing with a small user base. All I’m saying is there’d be better quality data to learn from if they try to mimic Uber rides. #tesla #autonomous #car #tech #teslamotors #tech #waymo #zoox #cruise #argo #argoai #nuro #baidu #aptiv #pony
Very few owners would sign up and they have plenty of data anyway why does it need to be through Tesla. The only way this helps is if you assume that then making it would change how people use their cars which is unlikely. Just sit back and wait for their share price to crash 90% to a reasonable valuation once people realize they’re not further ahead in FSD
Their teams, as per Andrej Karpathy are literally throwing neural nets at every edge case they identify. Ultimately they’ll address enough of those to get a general driving system (unlike localized solutions from the likes of waymo), just a matter of time. I doubt FSD will fail or the stock will crash.
Are you Cathy Wood?
Just an enthusiast, wanting to see someone crack Lvl5
Tesla should continue to take advantage of the FSD robotaxi believers until the bubble pops. People already paid 10k for nothing what's a little more.
How goes the HD mapping?
First video out of FSD 9 in SF and it's driving in the bus lane. If only it had a map 😂
Yes. And they might very well do it. The challenge I see is compliance and legal matters.