here we go again with the "Tesla Bot killers "
"the competition is coming"
in the words of SMR all over themselvesBMW maybe a compelling EV would be a good start
-
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
"the competition is coming"
in the words of SMR all over themselvesBMW maybe a compelling EV would be a good start
AnbaricBike
Member
Sorry, I guess I'll HODL my stories to myself, like I HODL my TSLA (since 2012!)I'm so glad that, while the stock market is open, VW diesel owners can trade stories and advice if they don't want to trade TSLA
No worries. You did excellent work at your Austin shop. Put a clutch in my Golf many years ago.
Knightshade
Well-Known Member
here we go again with the "Tesla Bot killers "
"the competition is coming"
BMW maybe a compelling EV would be a good start
To be clear, this is not "BMW to start making robots"
BMW is the customer of a humanoid robot company-- deploying those robots in one of their car factories
From the link said:
I love it!
One of the FUDdy angles used against Optimus is that it's just another far out Tesla/Elon-idea that will never amount to anything.
Having BMW and Figure doing exactly what Tesla and Elon has proposed validate this as a profitable business idea in the simple minds of Wall Street.
Ok. Well - perhaps in a few of them. One can hope.
It's similar to that NACS plug you guys use in the US. Lots of talking it down. Until Ford validated the idea by joining. Forcing others to join.
Now BMW and Figure has joined Tesla in thinking humanoid robots can make cars.
Only question left is who is next?
Sacconaghi at Bernstein best stock idea for 2024 was shorting tesla.
That is really annoying, and regrettably he has been right so far.
Compressed margins, limited production growth in 2024, and no
fsd yet. That is the underlying theme for now.
given the high p/e, we are vulnerable now
That is really annoying, and regrettably he has been right so far.
Compressed margins, limited production growth in 2024, and no
fsd yet. That is the underlying theme for now.
given the high p/e, we are vulnerable now
Last edited:
Interesting report out of China saying that Tesla had to buy cathode coils from tier 2 suppliers in China to bridge a shortfall of 4680 production. CATL and BYD declined to supply them since they only want to supply complete cells. Tesla hasn’t been able to make enough cathodes for their 4680 production goals, apparently due to their inability to yield cathode DBE. This occurred in 2023 and may persist well into this year.
cnevpost.com
Tesla importing cell parts from China to expand 4680 battery capacity, report says
Tesla began purchasing cathode coils from two second-tier Chinese battery makers in the second half of 2023 and shipping them to its Texas factory in the US to produce 4680 batteries, according to local media.
cnevpost.com
"Brighter with Herbert" -YT-channel has already a discussion about BMW/Figure -partnershio.
www.youtube.com
BREAKING: BMW Taps Humanoid Startup Figure to Take on Tesla's Optimus Robot
Get Free TESLA Milestone Tables My website: https://www.herbertong.comCheck out 15+ modules of resources for the $TSLA InvestorBrian White @4kpodcast (host o...
www.youtube.com
All EVs and autos in general are getting crushed. Rivian is down almost 6% today and Ford is down similar to Tsla.
Well, we know it can make coffee at least. Go Figure!
"10 Hrs Training Time" to learn the coffee trick? We need a new robot definition of Training Time (TT), Proficiency, and Efficiency; Imitation vs Reinforcement especially. And does the 10 Hrs include processing time? Or just what HR would put in as Training Headcount requiring a human to peer train only, then process the data offline for a week? Don't know, but 10 hrs complete seems unlikely.
However, if Learning Efficiency really is this good, we will see a robot folding a shirt within days. However, if it's imitating for 10Hrs then NN takes 2 weeks, we need to understand and clarify these metrics - or else we're back to... what is too cold? (This could take us to Elon's Jan 30th presentation for the real demo.)
Meanwhile, Reinforcement learning may take zero TT because it's exploring the feedback on it's own. This takes longer, but was TT = zero?
And when we talk about processing time, we need a metric there as well.
All these numbers can then be used to calculate True TT (TTT), for apples to apples comparison, and will be essential to understand total cost savings. From having a new training standard, one could extrapolate using a Task Analysis x Complexity of the Task at the atomic level. Math could then be used to estimate the robot ramp time.
Revised Terms:
TT = Training Time - Pure Human involved time (if any).
TTT = True TT - Includes NN Processing Time to Proficiency.
Efficiency = There's a formula already for this but needs to be re-written for Robots.
I had the fortunate opportunity to attend many workshops with Author Dr. Ruth Clark as my instructor, in person. One was specific to Learning Efficiency and how to calculated it. Safe to say, it's outdated where robots are concerned.
However, if Learning Efficiency really is this good, we will see a robot folding a shirt within days. However, if it's imitating for 10Hrs then NN takes 2 weeks, we need to understand and clarify these metrics - or else we're back to... what is too cold? (This could take us to Elon's Jan 30th presentation for the real demo.)
Meanwhile, Reinforcement learning may take zero TT because it's exploring the feedback on it's own. This takes longer, but was TT = zero?
And when we talk about processing time, we need a metric there as well.
All these numbers can then be used to calculate True TT (TTT), for apples to apples comparison, and will be essential to understand total cost savings. From having a new training standard, one could extrapolate using a Task Analysis x Complexity of the Task at the atomic level. Math could then be used to estimate the robot ramp time.
Revised Terms:
TT = Training Time - Pure Human involved time (if any).
TTT = True TT - Includes NN Processing Time to Proficiency.
Efficiency = There's a formula already for this but needs to be re-written for Robots.
I had the fortunate opportunity to attend many workshops with Author Dr. Ruth Clark as my instructor, in person. One was specific to Learning Efficiency and how to calculated it. Safe to say, it's outdated where robots are concerned.
Is this part of Tesla's long range plan to have all MYs and M3s with 4680s and giga castings?Interesting report out of China saying that Tesla had to buy cathode coils from tier 2 suppliers in China to bridge a shortfall of 4680 production. CATL and BYD declined to supply them since they only want to supply complete cells. Tesla hasn’t been able to make enough cathodes for their 4680 production goals, apparently due to their inability to yield cathode DBE. This occurred in 2023 and may persist well into this year.
Tesla importing cell parts from China to expand 4680 battery capacity, report says
Tesla began purchasing cathode coils from two second-tier Chinese battery makers in the second half of 2023 and shipping them to its Texas factory in the US to produce 4680 batteries, according to local media.
Doubtful. They need all the cells they can get and the 3 and Y lines are working fine with their stamped body sections.
Captkerosene
Member
Bot will be the Apple of robots; defined by great hardware, software and service. Premium products with premium margins. Even bigger TAM. Apple is worth $3T today.
It's a bold statement to say that Tesla will dominate in robots/automation. But look at Elon's record in developing other products. He's won in auto, space and it's starting to look like a win in media. In addition, Tesla already has the perfect infrastructure and skill set to develop and scale Bot. Bot can piggyback on compute, AI, Grok, inference chip, vision, OTA, marketing, sales, manufacturing equipment design and production, finance, engineering, HR/recruiting, distribution, parts storage and distribution, mobile service, factory design and construction and importantly the supplier network. Not gonna be easy for anybody to catch us when we're the fastest horse in the race.
But how fast can we ramp? Can we suck all the oxygen out of the room? The market will grow like other S curve markets. It'll be a challenge. But, a lot of Tesla's other initiatives are currently reaching completion. The new development in CT, M2, R2, Semi, Lithium mining, SC net, 4680 production and grid storage is mostly done. Lots of cut and paste to do but the hardest part is over. A fully focused Tesla could ramp production at an incredible rate.
Once scaled? Defend market share with continued innovation/updates. Apple or MSFT model work fine.
My estimate of present value for Bot/Dojo? (I do napkin math with binary outcomes.) Bad outcome is a zero or close to it. Good outcome is a 5T valuation for Bot/Dojo in ten years. 50/50 chance for each. Take the average (2.5T) and discount it back (50%) and we get a current value for Bot of $1.25T.
Importing cathode coils is not optimal as it is expensive, must be air freighted (if coming from China), etc. So if this report is true, then Tesla was probably facing a hard economic decision. Either idle/throttle their CT production line for low 4680 volumes, or import cathodes. Both cost Tesla money. This alone could explain $TSLA price action as their margins will/has taken a beating until this situation gets fixed.Interesting report out of China saying that Tesla had to buy cathode coils from tier 2 suppliers in China to bridge a shortfall of 4680 production. CATL and BYD declined to supply them since they only want to supply complete cells. Tesla hasn’t been able to make enough cathodes for their 4680 production goals, apparently due to their inability to yield cathode DBE. This occurred in 2023 and may persist well into this year.
Tesla importing cell parts from China to expand 4680 battery capacity, report says
Tesla began purchasing cathode coils from two second-tier Chinese battery makers in the second half of 2023 and shipping them to its Texas factory in the US to produce 4680 batteries, according to local media.
Then this should be top priority on the conference call. I don't want to hear Elon talking about f***ing robots.Interesting report out of China saying that Tesla had to buy cathode coils from tier 2 suppliers in China to bridge a shortfall of 4680 production. CATL and BYD declined to supply them since they only want to supply complete cells. Tesla hasn’t been able to make enough cathodes for their 4680 production goals, apparently due to their inability to yield cathode DBE. This occurred in 2023 and may persist well into this year.
Tesla importing cell parts from China to expand 4680 battery capacity, report says
Tesla began purchasing cathode coils from two second-tier Chinese battery makers in the second half of 2023 and shipping them to its Texas factory in the US to produce 4680 batteries, according to local media.
... what if that's the path needed to prevent TSLA SP from cratering further? Some investors feel the Robot is critical, myself included.
But agree, no more hype, just show me! Hopefully in actual use, including a projection, and not just folding shirts. (Unless it can also pick it out of a trash can, then go find a table to fold it on by itself.) Need some small proof of general intelligence applied to Tesla Mfg with measurement data described earlier (TTT).
I suggest you don't listen then.
GhostSkater
Member
With those 4680 news, I remember Drew saying the cathode plant would start dry commissioning by the end of 2023, but I don't think we've seen anything that supports that this has happened yet since there is still plenty of constructions and equipment being installed there
Might be that indeed current Cybertruck packs are not what they should be and initial owners are paying more for less in some aspects
Might be that indeed current Cybertruck packs are not what they should be and initial owners are paying more for less in some aspects
hallucinating about robots at this point will not have a positive impact on stock valuation
Similar threads
- Locked
- Locked
