Nvidia gives money to OpenAI so they can buy GPUs that don't exist yet with memory that doesn't exist yet so they can plug them into their datacenters that don't exist yet powered by infrastructure that doesn't exist yet so they can all make profit that is mathematically impossible at this point - Stolen from someone else.
quote from Tesla latest earnings call, at 04 min..
"Because we're really moving into a future that is based on autonomy and so if you're interested in buying a Model S and X, now would be the time to order it, because we expect to wind down S and X production in next quarter and basically stop production of Model S and X next quarter. We'll obviously continue to support the Model S and X programs for as long as people have the vehicles, but we're gonna take the Model S and X production space in our Fremont factory and convert that into an Optimus factory, which will... with the long-term goal of having 1 million units a year of Optimus robots in the current S/X space in Fremont."
I genuinely don't understand, is the Optimus real? Isn’t there a like 10 to 1 ratio of Boston Robotics demos to Optimus demos? Has it ever been verified to actually do anything?
Boston Robotics robots are over there doing backflips and the only thing I’ve seen Optimus do is in extremely controlled environments.
I second this. Is there anyone who actually believes Optimus is going to be a success and has any sort of data to back that up?
I'm not in robotics, but I look at humanoid robots and, while incredible examples of engineering know-how, they seem to be a long way from useful in commercial applications. Am I jhust ignorant of their true value? Seems like all I ever see them doing is parkour.
Optimus could do really well if they had all the smartest robotics engineers working on it...
But it seems that ~80% of the smart people I know refuse to work for Musk on principle, and the remaining 20% prefer to work somewhere that pays well (Musk companies do not).
End result is he has a team of mediocre engineers working on it which is why their demos appear years behind some competitors like Boston dynamics and Unitree.
I think the same is happening to Tesla cars (not much innovation in the last few years).
What has Tesla accomplished lately? I mean, within the last decade?
They certainly have accomplished amazing things. They had a lead that even five years ago was considered insurmountable. But they've made at best incremental progress, the kind made by mediocre engineers. The only novelty was the Cybertruck, which didn't live up to expectations and didn't open up any new domains.
SpaceX is still advancing, though even that is getting a bit of an asterisk if they can't get Starship to fulfill its promise.
I suspect they had an amazing team, but the last few years the best people have been departing and they are being replaced by mediocre people due to Musks involvement in politics.
Elon's hype level over Optimus practically off the charts. He has profit projections that have Optimus be effectively all of GDP in the future. Say what you want about Elon, but he does put his money where his mouth is and I believe he will try to manufacture robots. Also, the S and X models are old and their market segment is heavily saturated at this point so it makes sense for Tesla to exit those model lines.
Optimus is also a bit of a "squirrel!" for the market that he likes to talk about whenever sales figures at Tesla start flagging. Meme stocks only work as long as people still believe in infinite exponential growth.
> I second this. Is there anyone who actually believes Optimus is going to be a success and has any sort of data to back that up?
If there was no competition (there is), and he met the price envelope he's talking about (Cybertruck suggests he won't), I can buy the idea that there's a market opportunity for a few tens of millions of humanoid robots which are even just 5% AI and 95% remote workers in VR headsets, just because this means you can get cheap 3rd-world wages running your "made in America*" factories.
But there is competition, and I don't think he'll meet his price target.
As for the AI: even when I forecast under the assumption of continued improvements of hardware and software, I see at least a ten year gap between any given level of self-driving car and a humanoid robot small enough and light enough to get into the driver's seat of a normal car and drive it to the same standard, and that's just for driving a car, not all the other things people like to imagine in a world where androids are good enough to generically replace human labour.
* insert any other nation as desired, it works in any place where wages are higher than the cheapest nation with reliable internet (modulo the TCO of the robot, which nobody knows yet).
This will lead to minds getting blown as all those "foreigners coming here and taking our jobs" whose deportation people demand, are now "working from home in a different country and still taking our jobs" and the US in particular will have to wrangle with how this is a first amendment issue because remote control is just speech isn't it?
And when we consider how the current AI boom seems to have come with a total lack of even the most basic security considerations in their usage, these robots, wether** AI or remote controlled, are absolutely going to get turned into Mr-Stabby-the-totally-deniable-assassin.
** pun not intended, but when I noticed the misspelling I decided the implications still worked as a joke.
There is some value in producing a lot of solid hardware, but nowhere even close to Tesla's absurd valuation.
I think they are perfectly capable of writing software to drive the robot - if Musk doesn't stick his head in like he did with LIDAR/FSD and impose some stupid requirement that handicaps the product.
But the whole shtick with Optimus is that they aren't writing software. It's supposed to be all LLM training so when you buy your robot you can give it orders like "do the dishes", "clean the gutters", "dig a backyard pool for me", or "build me another Optimus" and you can go off to do whatever while it completes the task.
Elon thinks it would be too expensive to have to write code for every task you might ask one of these to do, they want it to be fully autonomous.
Their engineers aren't behind keyboards typing C++, they're wearing VR headsets and feeding the data to a LLM, although even that is probably too specific for Elon's long term plans. Obviously he doesn't want to have to have people repeat actions hundreds of times before the dumb robots figure it out. Especially for "simple" tasks like serving drinks at press events.
I feel like society is decades away from being comfortable with "you can go off to do whatever while it completes the task"...regardless of whether or not the tech is there.
It's just the AI singularity discussion again. AI Techbros insist it will be here before the end of the decade. Like you I am skeptical about it. I tend to think AI capability is already plateauing and ever more effort is going to be spent chasing smaller and smaller returns.
I'm experiencing AI that is very fast, but also kinda dumb and thoughtless.
You say this like it's a bad idea. These VLA models are going to be even more disruptive than the coding models because otherwise it's prohibitively expensive to set up an industrial robot for most uses.
My main doubts about Tesla's plan are that they will sell enough of these to get benefits of scale or that Musk will force the engineering team to "skip lidar" again and compromise the design.
> My main doubts about Tesla's plan are that they will sell enough of these to get benefits of scale or that Musk will force the engineering team to "skip lidar" again and compromise the design.
Indeed.
Even with 9 million total cars sold, Tesla still has yet to solve for driving safely with no interventions across just the contiguous USA.
With a similar approach, a million robots operating for years is still a long way short of gathering the data needed for training an AI to autonomously operate safely in a full range of industrial environments.
(That said, IMO remote-controlled humanoid robots still make a lot of sense, they'd only need a little bit of AI to assist rather than to do everything; if I was in Musk's position, I would be selling that vision of the future rather than claiming fully autonomous AI-driven androids are anywhere near).
It's yet another gamble where if it works out he will look like a visionary and if it doesn't he'll look like an idiot. The exact sort of bet that Elon never fails to go for.
But how would we evaluate "perfectly capable" without evidence, there's just been no evidence they've done anything so far right? Am I missing something? I guess looking closer it was only announced four years ago. But it seems like it's only been smoke and mirrors so far.
I think FSD is very impressive, even if it is still pretty unsafe.
Tesla clearly has at least some AI chops, and if Musk can bullshit for long enough, they might have enough time to make these robots more than just props.
And China is likely to do to Tesla robots what they’ve done to the cars. I assume the bans will be incoming, because the US can’t have millions of Chinese kung fu robots sitting about pouring tea, waiting for critical mass.
Optimus is a longer horizon promise that allows Elon to keep kicking the "can of untold profits" down the road. Tesla car hype has fizzled, robotaxi is currently fizzling, so the new promise is optimus. Elon sells dreams and visions, not really products.
Tesla absolutely cannot keep it's valuation without a promise for it's delusional stock holders or actual massive revenue streams.
This it could be the real strategy. Because the more credible promises you make, the more valuable is your company. If sales of cars are spiraling down, then what promises remain there to keep valuation ?
I think that the point of the comment was not that he does not sell any products, but that he predominantly sells dreams & visions, if you use TSLA market cap as a guide. If you look strictly at the products he sells, the valuation of his company ought to be somewhere between 50% and 100% of Ford. By that analysis it seems like TSLA is about 97% hopes and dreams.
You are correct to be suspicious, but don't be impressed by backflips. Those are just for show. Doing "real work" is the test. As is doing real work for a compelling price.
Yes, because they make money selling the parts, and there are warranty requirements that are hard to fulfill if you don't have parts.
Often after a decade or so, companies will sell the designs to dedicated parts makers. For example, Volvo has Volvo Classic Parts, and they even have a reman program, and will even 3D print parts not available. Mercedes has Mercedes Classic Parts. Chrysler has MOPAR, etc.
If you are a business, the costs of designing the part has already been paid, if you can sell the design and get some royalty payments, why wouldn't you turn those old plans into cash?
And of course there is a huge industry of Chinese clones and other suppliers that will provide replacement parts that are not genuine.
This reminds me that in the early days of Tesla they were complaining about the difficulty of competing on pricing with established automakers because they subsidize the cost of the vehicle at sale with profits from selling parts/service - a stream of revenue unavailable to a startup
It’s still possible to order new and original parts for SAAB models, almost 20 years after they went under. The spare parts are made by a separate company which is still going.
IIRC, by law manufacturers are required to maintain parts and service for vehicles for a minimum of 10 years. Whether superseded, discontinued, whatever.
Do you actually look at the current US landscape and think “the law is just the law” for the rich and poor alike?
Getting a judge to rule on something is also part of that “the law is just the law” and it’s obvious that judges are more willing to rule on cases for the poor and powerless than the rich and connected.
The point I was trying to make was that whether or not something is illegal is typically decided by a judge.
Most of the things corporations do aren't as clear cut as a traffic violation. So in those cases, you only know if something is illegal when it's made it through the courts
This is an urban legend. Safety defects have to be remedied by the manufacturer for a period of 10 years, but that remedy doesn't have to involve replacement parts.
I agree, looks like you are correct. It seems that it is just one of those things that manufacturers have agreed to do voluntarily, in the absence of a specific law. I imagine they have calculated that the loss of goodwill from abandoning a product quickly would outweigh the cost savings (especially since there is so much sharing of parts that keeping a few specialty components on hand is not going to move the needle much).
Yes. Auto manufacturers tend to have contracts with different tiered automotive suppliers that have heavy-hitting production lines for current vehicles, and also maintain a 'service' department where these style of products are produced. The tools for producing these parts have really good lifetimes, and you can take the tooling and put it into whatever mold machine you have written the program for, or set it up for another machine.
In my experience service departments are basically a large warehouse with a small set of assembly machines running at any given time where you are setting up time to produce some random part for a day or two and then change to something else, whereas the real production assembly lines are designed to produce as many of X part for the latest car as possible.
Several of the old mold machines where I worked that made parts for this service business ran DOS, with PCMCIA cards to load programs. I helped a process engineer get these PCMCIA cards working on his contraband laptop running win98 (obviously banned from the network) because we could never get them working with anything newer. This was in like 2021.
It depends. Lots of parts are shared by multiple models or even companies so it may be the case that nobody has made for example a new water pump specifically for your car for 10 years, but the design is the same as the 2025 something else so you can just use that one. There are also warehouses with older parts that can last for years. You can also pull replacements from junked cars that have not been crushed yet. In some cases third parties manufacture replacement parts when the supply runs out, but those replacements are often of poor quality and sometimes are only vaguely shaped correctly and require extra work to actually fit on your vehicle. Keeping old cars running is a challenge, especially if the car was obscure when it was new.
For traditional vehicles, there's typically a large marketplace of first-party and third-party auto parts for vehicles going back several years. Depends on the make and model, but usually yes.
That said, Tesla is a very unusual automaker in most senses and I'm not sure what their aftermarket parts situation is.
This is a concern for me not only for the Tesla but for the new Chinese manufacturers. When I've talked to owners of these cars (in other countries), the consensus seems to be "you use it for 5 years and then throw it away". Not because the car has poor build quality, but because there aren't local mechanics that can service it, it's impossible to find documentation such as torque specs and service procedures for anything but trivial stuff you'd find in an owner's manual, and it is very hard to find parts.
It seems like an incredible waste to throw away a car after 5 years.
A big part of what I look for in a car is a long lasting manufacturer that publishes to end users technical and repair information, including part numbers and procedures, together with a healthy third party part supplier ecosystem and independent repair infrastructure.
That doesn't mean that information needs to be available for free or that the parts themselves are cheap -- Volvo parts are not cheap -- but they are available and the information, engine specifications, repair manuals and workshop manuals are available.
If you don't have that, I'm not interested in buying the car. A car is far too expensive to treat as a disposable consumer good. I'm worried that more and more, manufacturers are locking down their systems, putting information behind paywalls where you can't make your own backup copy, and doing things like adding DRM to their parts to prevent indy shops from working on them.
What is this supposed to mean in the sense of Tesla being in the car game for the long haul? They're stopping production of 2 of their oldest model cars, but they're continuing to make and sell their most popular models globally the 3 and the Y.
The model S and X are their oldest most expensive models that aren't even sold everywhere in the world.
However the Model 3 and the Model Y are, of which the Model Y was the most popular car model sold globally of any manufacturer.
In Australia I see Model 3s and Model Ys everywhere. Stopping production of the S and the X, you wouldn't even notice it here.
They are going to have a hard time keeping it that way in 2026, since they were just barely ahead of both #2 and #3 (Toyota Corolla and Toyota RAV4), and as far as I know Toyota hasn't done anything to annoy a large fraction of the demographics that have bought the most of their cars in recent years.
This is the reason. IBM Mainframe business grew 60%. The modern mainframe is the best state of the art platform for computing, in both reliability and efficiency.
Also, IBM mainframes are wonderfully isolated from physical hardware. They could change processors in the next model, and users would notice a small delay as binaries were recompiled on-the-fly the first time it was used.
They surely could extract more performance from the hardware by shedding layers, but prioritized stability and compatibility.
> Also, IBM mainframes are wonderfully isolated from physical hardware. They could change processors in the next model, and users would notice a small delay as binaries were recompiled on-the-fly the first time it was used.
This was with AS/400's move from their own CISC processors to POWER. While you could pull that off with mainframes, it'd be recompiling actual native binary code. IBM mainframe architecture is very well defined and documented (sadly, unlike AS/400).
At this point in time it can have more cores and more memory than a Z, and likely higher performance in benchmarks, but the architecture is closer to a minicomputer than a mainframe.
"Charles Kushner is an American real-estate developer, diplomat, and disbarred attorney serving since 2025 as the United States ambassador to France and Monaco...
...In 2005, Kushner was convicted of illegal campaign contributions, tax evasion, and witness tampering, and was sentenced to two years' imprisonment, which he served in the Federal Prison Camp, Montgomery. As a convicted felon, he was also disbarred in three states. He later received a pardon issued by his son's father-in-law, President Trump, on December 23, 2020. Kushner has donated significant amounts to Trump's campaigns..."
Data centers consume...a lot...of water by design, recirculated water, does not means no water consumption.
Water must be continuously added in evaporative cooling systems used by many data centers.
[1] - Cooling towers reject heat through evaporation, which uses water, not just recirculates it. Evaporated water is lost to the atmosphere and must be replaced with "make-up" water. As a result, recirculating cooling loops still require new water input to make up evaporation and blowdown losses.
reply