-
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
I'm pretty sure they were real robots. A taste of our future, Optimus building more Optimus!
Really want to see more about the modular construction of the next-gen vehicle(s). Building modules of the left and right sides (with all doors and interior details) and then joining them to some form of center-vehicle module(s) is a radical approach. Really wonder just how they are going to get the structural integrity and join those modules into a completed vehicle. Concept makes a lot of sense, gotta be a lot easier to construct the sub-assemblies and Tesla excels on execution of new manufacturing technologies. Would really like to see that plant in operation.
One thing I've always been curious about with Tesla's constant design improvements cut in at random times is service/maintenance items. Configuration management has to be a nightmare to manage service parts. It's not like you can go into an Autozone and tell them you want a window actuator for a '22 MY (for example, not sure that particular part was changed) when it might be a different part through the model year. And depending on which plant it was built at. Seems like they are going to have to track right down to the VIN of the particular car, and keep lots of parts in inventory to support service.
One thing I've always been curious about with Tesla's constant design improvements cut in at random times is service/maintenance items. Configuration management has to be a nightmare to manage service parts. It's not like you can go into an Autozone and tell them you want a window actuator for a '22 MY (for example, not sure that particular part was changed) when it might be a different part through the model year. And depending on which plant it was built at. Seems like they are going to have to track right down to the VIN of the particular car, and keep lots of parts in inventory to support service.
Bored because you were a Software Engineer. You should really hang around Factory people more. Way more fun. We get out more, move more, have real tools, and make things you can touch, see, and even ride!
ChatGPT is a very useful component of AGI. It needs multi modal training (Ie add in vision and simulated real world physics with that vision), and the ability to manipulate (ie add more than text output, allow it to operate something like Teslabot), and the final piece of layering in short term memory into the NN, and you’ve got a Teslabot brain.
zoomer0056
Active Member
Well I'm glad we have a term to replace the over used "refresh" term. => Nex-gen
JRP3
Hyperactive Member
I don't remember hearing about a rare earth free permanent magnet motor from them before today.
This is literally the 30 mins I watched on YouTube in my car while waiting for my daughter.
I was impressed and jumped on here, hoping to find out what the rest of it was about.
Looks like I will need to watch the video
Some people are only impressed when the share price goes up.
What part of taking thin slices off a carrot is hard to understand?
It is a staged demo, not CGI. And the robots aren’t doing anything actually useful yet (Hence staged). They obviously need work in basic motor control (movements are very jerky), and they have a long way to go in AI control.
The picture is goofy, but I think it’s a genuine niche some companies can fill.
From the sounds of it, Tesla is not going to pursue every niche, they are going to crank out inexpensive vehicles in massive volume. Some people are going to hate that.
The big question in my head is how much of a premium are people going to be willing to pay for a more unique looking vehicle. Or for that matter a niche vehicle like the Jeep Wrangler. How big of a premium will people pay for a Wrangler?
It pushes traditional manufacturers into more expensive niches, but it leaves room for them assuming they are able to figure this out and produce these niche vehicles quickly. GM pursuing their Equinox is likely doomed. Tesla will crush them on low cost vehicles. But maybe the Lightning can survive if they chase the work-truck market.
I think a focus on the AI side Of Optimus would have been better tbh. In terms of robototic movement, it’s a big step forward from the previous display but very very simplistic compared to what a company like Boston Dynamics is doing these days
That’s the impression I got too.
I think that’s pretty impressive. It’s obviously still years out, but further than I thought they would be. Maybe next year AI Day they can have a demo like this on stage.
SageBrush
REJECT Fascism
My wife is not a Tesla nerd. The impression she wanted to share was that presenters were somber. She might have even thought unhappy. "Bad vibe"
<<shrug>>
My takeaways were different. I thought the overarching message was that Tesla is dead serious about producing 20M vehicles a year by ~ 2030 in addition to a TWh of storage. Music to my ears. This is why I invested in Tesla, and if the stock price drops before it rises, so be it. Actually, I embrace it.
Second, a basic truth remains that battery production growth is the linchpin of it all. So I listened intently to Drew and heard 'we are improving production by 1,000 units a week every quarter." I took 'unit' to mean enough cells for one vehicle. My spreadsheet tells me that works out 2M vehicles run rate in 10 years --- no where close. In fact, if my math is correct, the growth rate of the growth rate has to be ~ 2.
Here is the spreadsheet. Please correct as needed. Doable ? I think so, but it is one hell of a challenge.
<<shrug>>
My takeaways were different. I thought the overarching message was that Tesla is dead serious about producing 20M vehicles a year by ~ 2030 in addition to a TWh of storage. Music to my ears. This is why I invested in Tesla, and if the stock price drops before it rises, so be it. Actually, I embrace it.
Second, a basic truth remains that battery production growth is the linchpin of it all. So I listened intently to Drew and heard 'we are improving production by 1,000 units a week every quarter." I took 'unit' to mean enough cells for one vehicle. My spreadsheet tells me that works out 2M vehicles run rate in 10 years --- no where close. In fact, if my math is correct, the growth rate of the growth rate has to be ~ 2.
Here is the spreadsheet. Please correct as needed. Doable ? I think so, but it is one hell of a challenge.
Sounded to me like they were ready. They announced Giga Mexico and said it will produce Gen 3. So clearly they have some rough outline how the factory design should be. And they say they will copypaste the factory, not iterate on it like they did with Model Y.
Let's say it takes 1 year to build the factory like in Shanghai. Say 2years because of bureaucracy. Well then imo I think we can expect gen3 in ~2years... And likely they will build it in Austin, Shanghai and Berlin also with a few months delay or even before Mexico.
They were not ready to show it, because it was not a product reveal event. And maybe the event will be closer to production than Semi and Cybertruck... But imo it seems like they are getting ready prepare for making the factories right about now...
JRP3
Hyperactive Member
On Robots, yes staged but amazing still. Long way to go. The equivalent to the 2016 FSD Video. To anyone still working, about 10 years and everyone retires!
Task Analysis:
- Unplug wire.
- Grasp a complex arm object.
- Pickup drill and nut (but did not insert nut in driver, yet).
- Use alignment fixture.
- Pulled cloth off picture.
I know some MEs that cant do that drill thing, let alone Software people.
Task Analysis:
- Unplug wire.
- Grasp a complex arm object.
- Pickup drill and nut (but did not insert nut in driver, yet).
- Use alignment fixture.
- Pulled cloth off picture.
I know some MEs that cant do that drill thing, let alone Software people.
Xepa777
Member
Pre-programmed behavior, so honestly not that impressive on an absolute basis. Given the short timeframe, glad the new Tesla bot 2.0 can walk now and do some of these motorized functions, but that hyperbole on the tweet just screams "please give me attention" lol.
Unless you need to drive somewhere when you’ve just had a blackout. While cool and certainly helpful at times, I think Elon sees the relative inefficiency of V2G as *the* solution when more purpose-built solutions exist and are better. It’s kludgy to want to use the car for this. But I also see that if it’s all you have, it’s better than nothing. I think I would prefer to keep the car as a stand-alone backup. Cybertruck will fit that bill (when I add CyberLandr) and I just added the RPM trunk refrigerator to my MY.
Similar threads
- Locked
- Locked
- Locked
