Cost inefficient jobs, like sorting trash into recycling.
Time to turn landfills into money.
-
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
Time to turn landfills into money.
Artful Dodger
"Neko no me"
Yeah, if only Teslabot was available in Fukushima in March 2011. Would have saved a lot of bother.
I'm certain Elon is already planning to send Optimus to service the HST when it is reboosted by SpaceX. I wonder how much O2, water, and food a pair of Astronauts consumes? With Optimus, they can supervise the gyro upgrades w/o leaving the comfort of their own Starlink terminal...
Cheers!
Been skimming some comments from the AI day. Reddit, twitter etc is toxic AF... So many people with no technical understanding who are so fast to be dismissive, so much negativity and general hate. Sometimes I wonder if these guys are bots or not, but I think mostly they are just sheep being herded by some bot owner with bad intents. It's sad to see. And so many communities that used to be people interested in X are now filled with people who don't even like X. Finance is filled with people who hate people in finance, politics is filled with people who hate politicians, and technology is filled with people who hate technology. Really sad times we live in. Wonder how long this forum will remain readable. The Autopilot/FSD/Optimus subforum here is often very toxic, any discussion goes very quickly into long negative dialogues that makes me wonder if the posters are bots or not...
Anyway, I guess what I have learned from skimming the web is that most people were very unimpressed by Optimus and are downright negative towards Tesla FSD. People in the field or fanboys are very impressed, maybe a bit too much. There is very little inbetween...
Anyway, I guess what I have learned from skimming the web is that most people were very unimpressed by Optimus and are downright negative towards Tesla FSD. People in the field or fanboys are very impressed, maybe a bit too much. There is very little inbetween...
Xepa777
Banned
You see me fighting a holy war on those threads? I’m TheSource hahahha. Normally I just ignore but AI day jazzed me the **** up.
I embrace all the downvotes and am just spitting technical facts left and right. With shameless snarkiness hahahaha.
What I will say about here is….there’s inherently polarization because tesla has a completely different approach than any other company. That’s threatening to the industry and jobs. And it’s also threatening to us if tesla is wrong (and you do need several types of sensors for redundancy) and the lawsuit implications that could mean for the company. Mix that with a Elon missing timelines on FSD the last half decade and a highly complex subject, sprinkled with the stupid math on robotaxis, and it quickly gets polarizing. Good thing is that tesla does not need FSD for future valuation. But it’s a great hope to have.
Because bots will be building bots.
I don’t think this is true.
Certainly lots of nonsense on Reddit and Twitter, but that does not reflect general public sentiment any more than TMC.
Reddit and twitter are full of tons of people who know absolutely nothing but have tons of opinions.
TMC is full of people with a fair bit of knowledge and probably give tesla more credit than it’s due.
Edit: even after years of being pro tesla I don’t really know anyone that I have a personal relationship with that likes tesla. Despite my best efforts it’s a bunch of “competition is coming and Elon is a dick”.
Sunlight
Member
It's really sad how social media has ruined social and political discourse in a decade.
How do we backtrack from this without 'totalitarian' controls and controlled speech?
How do we backtrack from this without 'totalitarian' controls and controlled speech?
My take on highlights of the event:
- Optimus actually already runs on FSD computer, with the same occupancy network, retrained with indoors environment. This validates the basic idea of the program, that Tesla FSD real world AI is transferable to general purpose robotics. This puts Tesla on the leading edge of building an actually working brain for AGI robotic slave labor, mind blowing possibilities.
- Their latest hardware is optimized for cost and large scale manufacturing. Being useful(while being cost effective) is much more impressive than being able to do back flips. This put them a few steps ahead of any other robotic companies. Remember, prototypes are easy, scale production is hard, no other humanoid general purpose robotic company has the manufacturing know how to do that.
- $20k cost after scaling. First delivery to general public in 3-5 years! And that starts two orders of magnitude impact on economical output over FSD! I can’t even compute that…
- FSD inner workings are impressive and cutting edge, these we already know.
- Dojo replace a quarter of GPUs early next year, this would dramatically increase their iteration speed at lower cost.
- No words about HW4 at all, wise choice, keep it quiet until it’s already in the cars(which I expect would be CT first).
- Lastly, Optimus would broaden Tesla’s mission to “Making the future awesome“, opening doors for all kinds of new distractions.(/s)
You haven't been paying attention. Why don't you start with watching every second of this without skipping:
and when you are done, i will show you 100+ others.
We need to step back and ask some fundamental questions.
1. How much is Tesla spending on R&D?
2. Does it seem like Tesla is solving problems the right way?
3. If solved, will these products/services be valuable?
4. At scale will the in house solution be cheaper than simply buying from an external supplier?
5. Does Tesla learn a lot attempting to solve problems?
Fundamentally if Tesla wasn't vertically integrated and didn't do leading edge R&D Tesla would be just another car company and the products would at best be similar to other car companies.
When we look at the total R&D spend, other car companies are spending a lot more and achieving a lot less.
I don't understand why investors complain that Tesla R&D isn't good value for money, or why they try to limit the projects, and suggest purchasing off the shelf solutions instead.
When buying something, Tesla learns nothing, pays the same price as everyone else, and is limited to what the vendor is able to provide.
Artful Dodger
"Neko no me"
Well this prediction I made on March 31, 2022 panned out:
... and here's the slide from AI Day 2022 (see the 6:09 time index in Rob Maurer's Supercut)
So, Teslabot will have a 2.3 KWh battery pack (equivalent to 1 human liver's worth of glycogen-stored energy), that's the 'first principles' way to spec out the 'bot. But what else does this slide show? The pack appears to be a 14s-2p configuration of 4680s (52V / 3.7 V/cell = 14 cells).
Which, at current energy density of 100 wh per cell would yield a 2.8 KWh pack. But, as per best practices with the NCA bty in your Tesla (and this is the fun part), if you use the middle 82% of the pack, that's 2.3 KWh usable bty capacity.
First Principles Thinking™ Tesla-style.
Chairs for the Longs!
I bet it'll be 2.557 kilowatt hours (rechargeable in 20 min @ 3C)
That's one human liver's worth of glycogen energy (about 2,000 calories) plus 10% extra capacity for charge buffering.
Optimus' should beat humans handily on energy efficiency though (similar to ICE vs EV), and the AI CPU uses about as much power as a human brain (100w).
Once trained properly, I expect Optium' will be about 3x as productive as a human laborer, and able to work 21.5 hrs per day. With vacations and time off, that one 'bot replaces 6 workers. With its productivity, it creates the value of 18.
... and here's the slide from AI Day 2022 (see the 6:09 time index in Rob Maurer's Supercut)
So, Teslabot will have a 2.3 KWh battery pack (equivalent to 1 human liver's worth of glycogen-stored energy), that's the 'first principles' way to spec out the 'bot. But what else does this slide show? The pack appears to be a 14s-2p configuration of 4680s (52V / 3.7 V/cell = 14 cells).
Which, at current energy density of 100 wh per cell would yield a 2.8 KWh pack. But, as per best practices with the NCA bty in your Tesla (and this is the fun part), if you use the middle 82% of the pack, that's 2.3 KWh usable bty capacity.
First Principles Thinking™ Tesla-style.
Chairs for the Longs!
Last edited:
petit_bateau
Active Member
He is a Conservative MP. That's all you need to know. The clue is literally in the name, a Con.This sums up the stupidity of modern western NIMBYism right here:
God forbid a barren field of mud have zero emission solar panels placed on it!
These people are unreconstructed fossils.
GM has Nikola's HTML supercomputer!
petit_bateau
Active Member
Looking at those actuator slides and looking at shots of Tesla factories they already know how to make those actuators. Frankly looking at every single bit of the bot itself they already know how to achieve the $20k and >30% GPM. I am slightly surprised by the 2.3kWh, and my guess is that many applications will either have a corded power input, or a backpack power extender. Ditto for tools and other add-ons, both HW and SW.
This presentation was outrageously brilliant. I'm in Europe so watched it after my breakfast and it was gobsmackingly good. I stopped only once for a tea and a pee, when the questions started. Some of the meh questions were even turned into good responses, Elon et al have learned. The systems engineer in me was grinning from ear to ear when he said "emergent". I was stonked when I saw the semantic approach they are now taking to some aspects; fantastic convergence going on. We are indeed watching generalised AI coming into view, and I think I've said that before wrt what it is taking to bring FSD to fruition.
The whole presentation was excellent and pretty much fully covered the necessary topic space. Products, tools, theories; HW, SW, data, training, design processes, architecture selection, optimisation goals, utility functions, people, use cases, costs. FSD, DoJo, compute, Optimus. Learning, feedback, and correction loops. All these were necessary and sufficient for the ostensible event purpose (recruiting) and the underlying purpose (inform investors and industry), and a further purpose (motivate team to deliver capstone at a milestone event). Two areas I did not see covered, likely because they are less relevant for this purpose are the input/output communication layers for interacting with humans (hellish difficuty), and the (trivial) one of the acquisition cost-benefit (like, a large chunk of humans are literally not worth employing in the near future, ouch).
I do not see any organisation in human history that is/has performed so well at such scale and is accelerating even as we watch. Tesla is evolving.
Fantastic - and deliberate - show of the breadth and depth of talent. Poach her, not a problem. Vacation him, not a problem. Sabbatical him, of course. Not dependent on Elon. No one in that team is irreplaceable, that alone reduces risk of prima donna behaviour and reduces weak points.
Brilliant selection of problems, solutions, pathways at all levels (including choosing to go for Optimus itself). Great flexibility and learning in approach, changing course, backtracking, adapting. Delivering.
High correlation with SpaceX mission.
I am unsure whether inequality will increase or decrease amongst humans. There are a lot of factors in play. My expectation is it will grow, that is the long run default. If it does grow then there is a distinct possibility that being at the top of the stack in the future will be highly correlated with being a Tesla shareholder now.
I am blown away.
This presentation was outrageously brilliant. I'm in Europe so watched it after my breakfast and it was gobsmackingly good. I stopped only once for a tea and a pee, when the questions started. Some of the meh questions were even turned into good responses, Elon et al have learned. The systems engineer in me was grinning from ear to ear when he said "emergent". I was stonked when I saw the semantic approach they are now taking to some aspects; fantastic convergence going on. We are indeed watching generalised AI coming into view, and I think I've said that before wrt what it is taking to bring FSD to fruition.
The whole presentation was excellent and pretty much fully covered the necessary topic space. Products, tools, theories; HW, SW, data, training, design processes, architecture selection, optimisation goals, utility functions, people, use cases, costs. FSD, DoJo, compute, Optimus. Learning, feedback, and correction loops. All these were necessary and sufficient for the ostensible event purpose (recruiting) and the underlying purpose (inform investors and industry), and a further purpose (motivate team to deliver capstone at a milestone event). Two areas I did not see covered, likely because they are less relevant for this purpose are the input/output communication layers for interacting with humans (hellish difficuty), and the (trivial) one of the acquisition cost-benefit (like, a large chunk of humans are literally not worth employing in the near future, ouch).
I do not see any organisation in human history that is/has performed so well at such scale and is accelerating even as we watch. Tesla is evolving.
Fantastic - and deliberate - show of the breadth and depth of talent. Poach her, not a problem. Vacation him, not a problem. Sabbatical him, of course. Not dependent on Elon. No one in that team is irreplaceable, that alone reduces risk of prima donna behaviour and reduces weak points.
Brilliant selection of problems, solutions, pathways at all levels (including choosing to go for Optimus itself). Great flexibility and learning in approach, changing course, backtracking, adapting. Delivering.
High correlation with SpaceX mission.
I am unsure whether inequality will increase or decrease amongst humans. There are a lot of factors in play. My expectation is it will grow, that is the long run default. If it does grow then there is a distinct possibility that being at the top of the stack in the future will be highly correlated with being a Tesla shareholder now.
I am blown away.
Last edited:
Berlin. Was going to be end of Q4, was then pushed to end of Q1 23. Still planning for 5000 cars/week at end of Q4 22.
Tesla pushes back plans for full third shift at Giga Berlin, but 5,000 cars per week still possible before end of 2022
Tesla opened Gigafactory Berlin in March 2022 and was able to ramp production quickly enough to add a second shift just two months later. The two shifts means new Model Y cars are rolling off [...]
driveteslacanada.ca
Rob put together a condensed (23 minute) supercut of the presentation.
I wonder if EM pushed this to Friday Sept 30 because he knew that from an SP perspective it will mean nothing on Monday after P and D tomorrow.
It served its purpose as a recruiting presentation without allowing the hedgies to attack relentlessly for days over the fact the Tesla lied and they don't have an army of robots to conquer the world.
I cannot claim any expertise in robotics, but I don't see how you cannot be impressed by how fast they produced a walking robot. And I thought the presentation itself was pretty cool.
And it seems FSD will steadily improve from here. Maybe it is only a year away. The current version would probably work fine if all cars on the road actually ran on FSD. Those pesky humans...
I am confident that a year from now Optimus will be very impressive. Will it be the best robot out there? Maybe, but I will definitely go on record and write that it will be the cheapest, and that might be the only thing that matters.
It served its purpose as a recruiting presentation without allowing the hedgies to attack relentlessly for days over the fact the Tesla lied and they don't have an army of robots to conquer the world.
I cannot claim any expertise in robotics, but I don't see how you cannot be impressed by how fast they produced a walking robot. And I thought the presentation itself was pretty cool.
And it seems FSD will steadily improve from here. Maybe it is only a year away. The current version would probably work fine if all cars on the road actually ran on FSD. Those pesky humans...
I am confident that a year from now Optimus will be very impressive. Will it be the best robot out there? Maybe, but I will definitely go on record and write that it will be the cheapest, and that might be the only thing that matters.
Nolimits
Member
I was truly impressed w the team actually. Looks like they have good talent working for them. Most of it was over my head but if I was an AI / robotics engineer, I would totally apply. As an ortho surgeon myself, I was intrigued by the actuators in the joints and wire/springs used in the hands to recreate flexor/extensor mechanisms. Curious to know how they designed the shoulder as it is a complex joint. I don’t think they went over that. I am sure OrthoSurg could chime in.
Also on twitter, people were saying they had to sign a waiver to ride in some vehicle at the show. Potentially dangerous and could die. Any videos or clips of that? Was it an autonomous cybertruck? Prototype robotaxi?
Also on twitter, people were saying they had to sign a waiver to ride in some vehicle at the show. Potentially dangerous and could die. Any videos or clips of that? Was it an autonomous cybertruck? Prototype robotaxi?
Similar threads
- Locked
- Locked
- Locked
