ThisStockGood
Still cruising my Model S 70 2015
He’s talking about the first split. Not this one.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Tesla, TSLA & the Investment World: the Perpetual Investors' Roundtable
- Thread starter AudubonB
- Start date
This is the standard fear-assuaging discourse from someone important when leaving a company. I’m just one cog in the machine, the team is great and will plow ahead without me, etc etc
Those are the words, actions will bear them out (or not)
The last four months already support them.
Right, we haven’t even had a split announcement in 2022 yet. The shareholder vote is step one to allow a split to happen, not the announcement of a split itself itself.
Last edited:
bkp_duke
Well-Known Member
Spot on. This guy has been out of actively AI development with Tesla for the past 4 months, while FSD has continued to march along with steady improvements.
lafrisbee
Active Member
Yay! (not really)
I never did get an answer to my semi day trading question when TSLA was at $720 (nor could I find the answer on the net), so I sat on my hands...
But I am sure I am free to play the pennies and steam rollers game now that it has dropped to where I thought it would be today... still rich.
I never did get an answer to my semi day trading question when TSLA was at $720 (nor could I find the answer on the net), so I sat on my hands...
But I am sure I am free to play the pennies and steam rollers game now that it has dropped to where I thought it would be today... still rich.
Right. Saying cruise cars are fully autonomous is splitting hairs. In practice, if communications with the mothership goes down, Cruise cars get stranded. That’s how how you can have twenty cruise cars sitting idle. All wireless communication systems go down from time to time in certain areas. So basing a system that requires it is pretty fragile and not scalable.
If your account has less than $25k in it - you cannot day trade more than (I forget the frequency) a couple of times per 90 days.
If you are trading in a retirement account - you have to wait for the funds to settle from your trades before using that money again - 3 days.
I would imagine it is one of those 2 instances happening.
Isn't this due to California regulations tho? All autonomous cars must have a backup remote control system. So I suspect each cruise car will be illegally operating when the backup controls go down so they are forced to stop in case of such event.
TheTalkingMule
Distributed Energy Enthusiast
This SP manipulation is an outrage! Oh wait......they're supporting TSLA today?
Long live the MM manipulators!
Long live the MM manipulators!
The whole market seems to have a lot of support here. We've had two bad news days in a row and it can't break lower.
I trade in my retirement accounts all the time. You can get margin approval, which isn't really margin, but let's you use funds that haven't settled.
Bad ass. That's got some great links for railway. I would bet Tesla is planning on running things nearly straight south and right up the the doorstep of the Austin Gigafactory.
For anyone that cares, here is a map of US railways:
Freight Rail Map of Class I Carriers in North America - ACW Railway Company
The Aberdeen Carolina & Western Railway Company runs from Charlotte, NC to Pinehurst and connects to Norfolk Southern and CSX Transportationwww.acwr.com
Lawrence KS --> Topeka KS --> Wichita KA --> straight through Oklahoma --> Ft Worth, TX --> Waco, TX --> Austin, TX
~700 miles, or less than one day for a loaded train
Freight on that should be dirt cheap, compared using trucks.
It is interesting that quite a few of us have been in closely analogous positions. Bluntly, the people who are best at leading a new technology development or even a startup of any flavor are usually (that is NOT always, BTW) not the ones who are best in running something that follows the first phase. There are endless examples in human activity, technological or not. In my own career I led and/or participated in numerous startups, turnarounds and new technology adoptions.Thanks for this.
I have a slightly different story. Many years ago, I left a major consulting firm after leading a small dev team for 5 years. I'm not one who seeks to keep climbing higher, but I do believe that innovation requires change. I felt, and so did my team and peers (anonymous peer review contributed to 6 years of "exceeding expectations"), that I'd led our terrific team to great results. But after 5 years, I suggested that management switch us managers around between teams (in particular, I felt the "Lotus Notes" team could use my innovation approach to become a true messaging team. I thought my team had accomplished as much as my skills could help them with, and new blood would help them grow in areas I wasn't as strong in (e.g. documentation, process optimization). Management wouldn't change. So I moved to a completely different department (but only after my boss insisted I help hire my replacement!)
I'm SURE there are super bright minds and leaders in Andrei's team. We all wish them every success, as did Andrei himself on the Twitter sign-off.
The turnaround cases I have just reviewed. All but one was to fix what founders had failed to do. The startups, all but one, I left soon after success was established. The departures were voluntary in all but one.
The one that was not is one I felt was entirely dependent me. I was wrong and I failed that time, quite abysmally. The successor organization made it into a major competitor in the field.
Bluntly and clearly, I applaud people such as J B Straubel, Karpathy and others who know when to depart. The gigantic exceptions prove the rule, people such as Elon Musk. Even they know how to both select people and nurture them while letting them leave gracefully when the time has come.
Personally I find this development to be positive. As with other stellar figures, the general public thinks their departure is a danger sign. Of course it is nothing of the kind. It is a sign of organizational maturity.
"iteration" usually is "tweak some parameters & recalculate". That is a problem you can throw compute at, IF you have good people tuning things. Kaparthy said, that most issues were with what you put into the mix in what percentage. I.e. tesla once had too few tunnel-images for their rain-sensor-NN.. the wipers went mad inside tunnels. Same for things like "yellow light", "cone moved in your lane by wind".
But that "raw compute" only gets you so far, as all it does is getting things nearer to the local optimum - and usually an order of magnitude more compute/tries per percent improvement.
That is not compute-intensive at all. At my last employer i wrote such an "in-the-loop"-system to get more trainings-data. The way way most expensive part is the human labor of annotating. And given you GOT such a team, then the big question is "what" to label. This ties into the previous point: What data is worth labeling to improve & make the massive compute you throw at each iteration worth it.
I was constrained by having only 2-4 "Labelers" doing this for 1h/day on 2-4 days/week. It was extracting structured information from job-ads. They managed to get ~100 Ads done per hour per labeler. To decide what was worth i devised a metric i called "confusion" that strongly corrolated with wrong answers from the net - without having to know the true answer. It would have been impossible for them to label all ~10k ads that were processed every day
Think of it like a driving-scene.. You label everything in 3d-space with 99% accuracy for the top-class, 40% for the second-best. Then you have one object there that has "only" 98% in the top-class, but 80% in the second-best class. Then you have a "confused" state, where 2 classes could be true. Real world sometimes is ambiguous. But imagine one thing being in "drivable" class (like painting of a stop-sign on the road), one being "non drivable" like a stop-sign painted on a brick wall.
These are exactly the examples you want to "discover" and "bring up" in your huge pile of data. Or - in the case of tesla - data that is not temporarily coherent. Like at the beginning of the clip you say "80% tree" but 3 seconds later "90% thing painted on van". This also is a discrepancy you want "fixed" in your data.
And with each iteration of the network it will be confused about other things. 2 steps forward, 1 step back.
Also after some time i started logging WHICH annotator did what. Because people are inconsistent. Sometimes they label the same thing differently when presented with the same data on different days (a problem that came up regarding data in my bachelors thesis -.-). Sometimes they think they understood what they should do - but don't. Then you have to throw away whole batches of data from that annotator. And go through the past data (or mark them for relabeling through others) and correct them as well.
In real-world-ML-applications you lose 90% of your hair over the data, annotation and preprocessing... not the computation, algorithms etc.. that is only the other 10%
tl;dr: Tesla laying off labelers but still not binnig FSD can be thought of as bullish, because each iteration needs less annotation to "fix errors".
Even if you have infinite compute the limiting factor then will still be annotators because humans are slow (and need to be supervised (random samples) for this -,-).. and programmers that tweak & fix things and i.e. mark stuff for annotation or select new data (like sending a request to the fleet to catch "wheel flying through the air" and similar).
fully agree.
The Accountant
Active Member
adding more details for @lafrisbee
If you open and close a position on the same day it is considered a Day Trade.
Merrill Lynch states "if you make four day trades in a five-day period, you're classified as a pattern day trader and subject to specific margin requirements.
You can find more info here: Day Trading rules at ML website
lafrisbee
Active Member
Thanks guys...
And Accountant, I'll go read it
Haven't read all the fine print but it seems like this is applicable for all states, a backup remote system. Maybe knighshade can chime in.
To give you an idea of how bad the natural gas crunch is going to be for Germany this winter, here’s a forecast study that assumes people will start resorting to burning wood as a replacement. This is de facto de-industrialization. Hopefully this will be fixed within a year. But it highlights very real risks for Tesla’s Berlin plant this winter since I assume it uses natural gas.
Knightshade
Well-Known Member
To my knowledge there is no federal regulation of self driving systems at all so not sure what you're seeing that you think requires remote backup in all states?
In the half a dozen or so US states that self driving (no special approval needed) is already legal I'm unaware of any such requirement at the state level either... (CA is NOT one of those states- they heavily regulate and require tons of approvals at various levels)
Similar threads
- Locked
- Locked
- Locked
