Pras
Member
Let's just say, Fred had it coming!
Recently, Fred didn't give Tesla any (as we say in golf) forgiveness.
Recently, Fred didn't give Tesla any (as we say in golf) forgiveness.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
-
Tomorrow at 1PM PDT we will be discussing the latest Tesla Shareholder Meeting on the TMC Podcast. You can watch live on YouTube or X and participate in the chat.
Tesla, TSLA & the Investment World: the Perpetual Investors' Roundtable
- Thread starter AudubonB
- Start date
Without investigation results, making such baseless claim would get him into another trouble. Not that he hasn't done that before.
Pras
Member
He is pissed as I am because yesterday the headline story in the morning on CNBC was tesla is on fire.. and it was spread all over. When he questioned this bias, Fred tried to defend the media and he responded what we all felt already about Fred.
I was watching cnbc's youtube clip about the fire and I can't believe they were defending Tesla, saying how they need to stop posting these videos because hundreds of thousands of fires happen a year and it's unfair to Tesla. I was floored.
anthonyj
Stonks
SpudLime
Active Member
I agree with your prediction, though I interpret it as a direction for the whole company and perhaps a more modest growth of R&D per se. Of course, costs will still be watched by the gimlet eye of Mr. Musk.
Perhaps we will be able to infer some further interesting business choices as they converge on autonomy. Elon mentioned the model mix (the advantage of making 3 SR+'s rather than 2 LR's) and also giving no option to keep the 3's at the end of a lease.
As both their understanding of their future business improves -- now with autonomy! -- and their confidence in their technology is validated, we can expect to see some interesting developments.
Before they throw the switch to enable the million Tesla's out there at that time to become robotaxi capable:
- How will they be selling to the people most likely to commit their vehicles to the network?
- Will us current owners be importuned to commit our oh-so-lovely vehicles to the network?
- How might Tesla favor nations/jurisdictions that open up to autonomy first?
- Might they take a cut from the insurers as they reduce the insurers payouts because the Tesla vehicles are demonstrably so much safer?
- oh, and other juicy somethings for later discussion...
- What if you have Boring tunnels and can thus save people travel time, what pricing power does that give you?
Etna
Member
Zhelko Dimic
Careful bull
Actually, hardest and least well understood part is infrastructure and tooling needed to develop such a massive system, under which I also roll scalability, deployments, CD/CI pipeline, unit tests etc, etc. This includes all auxiliary things people don't think about: systems for collection, review, tagging of data from the car, communicating triggers to the fleet, processes to provision and stand up training infrastructure, feed it data, dismantle, etc, etc. All the shadow modes and configs that no-one ever sees, but are mandatory for proper evaluation and data driven decision making...
I was impressed with what Tesla has so far, you could see glimpses of it from their simulator, or debugging tool that lets them render 3D scene of data that car sees (from any vantage point). That means they can roll through time and space, effectively analyzing and coding heuristics of actual driving. Third guy presentation was most impressive to me, as someone who's done software for 25 years. They're in a good spot there. FSD will continue improving relatively rapidly.
Having said that, I share your concern, that they also voiced (I think Karpathy, but maybe third guy), that hand coding heuristics for every additional case will plateau. Sure, they can find things that once well understood could be pushed down to NN, downsizing code complexity they're dealing with, but... I don't know how they can heuristically solve the problem of right (or left) turn in downtown Manhattan on unregulated crossroad where pedestrians keep coming. Or any other situation that requires actual negotiation (through actions) between actors and involves taking calculated risks, counting on other party to abort their actions. Like inching forward, to severe flow of pedestrian traffic, until you can move... Those things that involve eye contact specifically. This involves many downtowns, plenty of European inner city driving, third world driving, some parking structures, temporary events spaces (as you pointed years ago) etc.
I mean, they're solvable by emulating human behaviours, but it will take plenty of time. This is why I got annoyed when Musk promised downtown NY driving by the end of the year (this or next? - doesn't matter) They'll solve for 95% and that could be a great business. To solve for last 5%, answers is maybe in further push of heuristics to NN, and if so, it will probably take another 2, maybe 3, maybe 5 orders of magnitude more complex NNs, so it's at least years away.
This event was great.
It would have been fantastic without Musk.
What a way to snatch defeat from the jaw of victory.
Last edited:
What the hell is wrong with the media? Laura Kolodny reporting today stinks. How pathetic can one be to be reporting like this, "Elon sent two line email to his employees on FSD demo, and promised 1 million Robo taxis next year"
Where did these guys come up with 1M Robo taxis in 2020?
Where did these guys come up with 1M Robo taxis in 2020?
Perhaps not. There was a thread long ago (like three weeks) by @generalenthu that analyzed FCA's problems for 2019 and I tried to assess the level to which pooling with Tesla could help them.
I know you are kidding, but if you are not - that is not a dig on Nissan or any automaker. He simply said he will make a quiet Leaf blower.
Last edited:
Pras
Member
No, I am talking about CNBC.com top article. they covered this news so many times and of course we longs were happy to praise the one little instance they covered it positively. Although, that video was Phil Labeau and not average CNCB view.
I wonder how much the Full Self Blowing package will be...
Yesterday's Autonomy Investor's Day put me in mind of the early tests of air power.
Can't find moving pictures of the event, but here's a photo of the first USS Indiana (BB-1) after the tests in 1920:
There's also the sinking of the Ostfriesland in 1921 and more in this video:
These are biplanes mind you, but this is the sort of thing that happens when you can add (and dominate) another dimension in a field of, uh, endeavor.
And, yet, many people who presumably saw the test results still didn't really get what they were seeing...
Can't find moving pictures of the event, but here's a photo of the first USS Indiana (BB-1) after the tests in 1920:
There's also the sinking of the Ostfriesland in 1921 and more in this video:
These are biplanes mind you, but this is the sort of thing that happens when you can add (and dominate) another dimension in a field of, uh, endeavor.
And, yet, many people who presumably saw the test results still didn't really get what they were seeing...
Last edited:
He is probably listening to a loud gas leaf blower right now and is annoyed. As am I when I hear them.
HG Wells
Martian Embassy
No SR+ inventory within 200 miles of Atlanta.
No model 3 inventory within 200 miles of Miami
No model 3 inventory within 200 miles of Miami
Similar threads
- Locked
- Locked
- Locked
