Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Tesla, TSLA & the Investment World: the Perpetual Investors' Roundtable

They want to sell it but there needs to be buyers. Tesla's credibility in this area has always been over promising and under delivering.

However if FSD 1.0 becomes a thing using DOJO then they will be showered with demand. Before that, Elon's credibility in this area is as good as "FSD L5 by the end of this year"...which you know I am hopeful for...
If ARK's projections in their Big Ideas 2021 come true, Tesla will have no problems finding customers for Dojo:

"While advances in hardware and software have been driving down AI training costs by 37% per year, the size of AI models is
growing much faster, 10x per year. As a result, total AI training costs continue to climb. We believe that state-of-the-art AI
training model costs are likely to increase 100-fold, from roughly $1 million today to more than $100 million by 2025."

ARK-deeplearning.png
 
Is this a known fact? Is there a source? I mean I knew it probably wasn't a dozen nerds in a basement but that sounds like .... a lot.

Many analysts seem to think FSD will have something 98% margin. That's gonna take several years to get to with that many on the payroll.
Source:
 

jw934

Member
Supporting Member
Jul 23, 2018
508
10,493
Toronto
GEE, I would at least wait till fall to announce it.
It snowed today across many states and provinces today. Would be a good day to have a heated steering wheel if Tesla made it available today.

WISN 12 News tweeted a horrific crash captured a few hours ago by "Tesla Cam". I am most impressed how the driver and the Tesla handled the slippery road to come to a safe stop despite originally going much too fast, seemingly following too close for the white-out condition, and having to dodge through the dynamic wreckage situation.
 
Just an after-hours thought:
Unlike some years ago I noticed that I am about the only left who who still has a (reasonably) realistic profile photo.
Is that because everybody else now has to protect their privacy after they became billionaires?
Either way, it's boring...

GEE, I thought that was a photo of Gerald Ford when he went biking there in the 1950's.
 

Chunky Jr.

Hungry Member
Supporting Member
Mar 9, 2018
2,064
32,047
CA
Good effort today by the home team. Fell behind early, but clawed their way back to a nice lead by halftime. They came out flat in the second half, but got a second wind in the 4th quarter to pull away for an easy win. These are the kinds of wins the loyal season ticket holders have come to expect, and it is nice to see them looking like their old selves the past couple weeks. Hopefully this is just the start of another solid run.

This is the first time they've been 3 games over .500 since March 1st.

Today
Score:
744.12
Margin of W/L: 25.13
Attendance: 31,146,428
High - Low: 46.84

Season
Record:
39-36
Total margin of wins: 901.89
Total margin of losses: -863.44
YTD gain/loss: 38.45 5.45%
Best Win: 110.58 Mar 9
Worst Loss: -68.83 Jan 11
Last 10: 7-3
Streak: W2
Avg margin of victory: 23.13
Avg margin of defeat: -23.98
Avg Attendance: 36,183,945
Avg Attendance of Last 10: 33,547,496
Avg High - Low: 41.69
Avg H - L of Last 10: 31.14
 


Indeed, some still cost more than others to insure, and Insurance.com has determined the most expensive cars to insure for 2021. The company calculated full coverage for a single 40-year-old male with a clean record, good credit, and 12 miles of commuting to work each day. This hypothetical driver has policy limits of $100,000 for injury liability for one person, $300,000 for all injuries, and $50,000 in property damage. He opted for comprehensive and uninsured motorist coverage as well as a $500 deductible.

View attachment 655831
Not sure I buy the figures presented by MotorTrend. Tesla's are quite inexpensive to insure.
My two vehicles which are identically insured covering my wife and I, are our 2018 Model 3 LR and 2008 Honda Pilot 8 seater, and the Tesla is $25 less expensive per annum to insure than our early model Honda.
 
Last edited:

BornToFly

Active Member
Supporting Member
May 8, 2013
3,251
26,499
43,000 feet
I would be excited for the FOMO price increase on Monday if the stock chart of today had just held off for one more day, but now I will have less enthusiasm to see tomorrow. but should the MM's are unable to manage a lower price at close tomorrow i will be again excited for the monday FOMO,
We need several up days in a row to keep raising the upper BB, otherwise a quick pop will be brought back down. Upper BB is only 774. Hopefully we get close to that tomorrow and it raises to 800 or more for a close on Friday around 800, making further room for a climb on Monday.
 

capster

Active Member
Supporting Member
Aug 11, 2018
1,203
10,001
An island planet
Just an after-hours thought:
Unlike some years ago I noticed that I am about the only left who who still has a (reasonably) realistic profile photo.
Is that because everybody else now has to protect their privacy after they became billionaires?
Either way, it's boring...
We’re all going to be doxxed in the end.

For the trolls‘ information (not meaning you @lklundin), should truly appreciable climate impacts happen sooner than later, the servants of oil (pronunciation below), including those on the Street and in the media, should probably worry.

 

capster

Active Member
Supporting Member
Aug 11, 2018
1,203
10,001
An island planet
Achieving FSD 1.0 has nothing to do with DOJO. DOJO is just a faster way to train the networks they are currently training using GPUs. (They could release FSD 1.0 before SOJO is up and running.)

DOJO is purely about computing power focused on AI training vs. using off the shelf GPUs. The only advantage of DOJO will be iterating faster, it isn't going to make FSD better just by switching the training from GPUs to DOJO. (Like some people think.)

Even if Tesla never solves FSD DOJO could prove to be of great utility/value to other people/companies.
In my experience in the land of software, iterating faster does lead to better results even though it may not seem like it should.

It’s one of those "in theory there’s no difference between theory and practice, but in practice there usually is" kinds of situations.
 
meaning there won't be much time available for external customers to use it?

We probably need to distinguish between DoJo hardware and software.

If there is only one copy of the hardware it may be used for Tesla internal use, but once the hardware and software platform is proven, making additional copies of the hardware is possible,

Vision based AI can be applied to lots of problems - e.g. processing recycled waste. geological surveys, exploration, mapping. astronomy, search and rescue

Any area where a human is doing a repetitive and menial tasks, that require vision, is a candidate.

Dangerous tasks like bomb disposal, land mine clearing clearing are others.

Finally we get to extreme environments like deep water, where drone submarines can be made smarter.
 
But that assumes that you are still working and making some new money. Who here does that???

Or that you are not already 100% in Tesla so you can undiversify. Who here isn't???
I'll still be working for the next 1 to 9 years, I guess if TSLA advances enough I might tell my boss to shove it one day.

If not I'll retire when those Required Minimum Distributions start hitting my checking account (especially if they are in the 6 figures or greater style, which currently happens around 4200 a share).
 

MP3Mike

Well-Known Member
Feb 1, 2016
19,422
48,270
Oregon
In my experience in the land of software, iterating faster does lead to better results even though it may not seem like it should.

It’s one of those "in theory there’s no difference between theory and practice, but in practice there usually is" kinds of situations.
My point was that the NN spit out of their current GPU based cluster would be identical to the NN spit out of DOJO. The only difference being that the DOJO one is ready for testing first. And obviously being able to train your NNs faster means you can iterate, and fail, faster. Which gets you to a better end product faster.
 

gabeincal

Active Member
Supporting Member
Jul 5, 2016
1,298
10,669
SF Bay, CA
In my experience in the land of software, iterating faster does lead to better results even though it may not seem like it should.

It’s one of those "in theory there’s no difference between theory and practice, but in practice there usually is" kinds of situations.

Kinda like a stock split? ☺️
 
I scrolled back a few pages and haven’t seen mention of this new ARK whitepaper on solar, storage, and crypto mining:


When I said a few months back that Tesla could literally create a brand new economy based on energy, this is what I was talking about. Another piece of the puzzle.
 

M|S|M MYP

Active Member
Supporting Member
Feb 20, 2018
3,891
38,364
USA

Products we're discussing on TMC...

About Us

Formed in 2006, Tesla Motors Club (TMC) was the first independent online Tesla community. Today it remains the largest and most dynamic community of Tesla enthusiasts. Learn more.

Do you value your experience at TMC? Consider becoming a Supporting Member of Tesla Motors Club. As a thank you for your contribution, you'll get nearly no ads in the Community and Groups sections. Additional perks are available depending on the level of contribution. Please visit the Account Upgrades page for more details.


SUPPORT TMC
Top