> Employees at Meta, the parent company of Facebook, Instagram, and WhatsApp, have expressed concern over what they describe as “obvious” and “biased” favoritism toward members of the Bolsonaro family during Brazilian elections.
Incredible, and here I was thinking the letter was going to be in regards to the last minute extension of the election by Lula's former attorney, or the irregularities in the second round. Facebook is a culturally left-wing company and their argument seems to lack substance
You say that but the economy seems pretty strong to me, we have lower than average unemployment with above average job listings. It's harder if you're on visa otherwise it's pretty competitive
As far as I'm aware Tesla has always been very generous with refunds for deposits, even for cars that were produced. For the model 3/y/etc they would let you cancel the deposit if you pre-ordered and just no longer felt like buying it, as long as it didn't enter production yet.
It wasn't pulled out of a hat, it was the total addressable market for a space, internet, and AI company, excluding markets like Russia and China. All TAM refers to is the total market size, not their market share.
For context SpaceX's TAM is lower than Anthropic's expected TAM
The business is doing fine, they’re profitable and regularly beating various targets. Honestly I think Elon is more level headed than people give him credit for, his companies are consistently profitable and generally produce competitive products.
Operating margin fell from 16% to ~2% over the last 10 years and will probably be negative in 2026 and 2027. They have a big cash pile to burn through but they’re spending it on ummm “a flying car,” humanoid robots, and an AI fab. They’re fucked.
Similar, to the time Tesla was fucked, because they wanted to sell EVs when there was neither demand nor charging infra. Or when SpaceX wanted to build a rocket that could land on legs, whereas every wannabe rocket scientists knew that to be impossible.
Yes, Elon's time tables are very aggressive. But that is by design. And FSD is here, although it took much longer than anticipated, no competitor caught up.
I'd be very surprised if the roadster project was particularly expensive. That operating margin tells you they are spending all the cash they are earning on cars on the other stuff, not that the cars suddenly became more expensive to make.
I'm interested to meet someone online who's short both on-shore US silicon fabs and the market for humanoid robots. Can you say more about your thinking?
I'm short on an allegedly successful automotive company that has a rapidly deteriorating automotive business (due entirely to management's own decision-making) pivoting into the most hyped business ideas in the world. I'm bearish on the type of creative accounting and self-dealing we see between the Musk companies.
Musk is similar to Zuck in that they get some credit for sheer courage to turn the wheel on their massive businesses, but they also have to be right. Aside from his anticompetitive behavior, Zuck continues to miss. Musk's pattern has been that the more Musk is personally involved, the worse the outcomes. Cybertruck was his baby. Optimus is his baby.
If Tesla is doing so well as an automotive business, then why doesn't he just continue making it a great automotive business?
I mean the short answer is that it doesn’t temperamentally suit him to do so.
Maybe a mid length answer is if you agree with his assessment on say humanoid robots in terms of market and you believe they might have a geometric growth factor in production (eg robots can make more robots) then you’d want that business build out done by a team that can scale rapidly.
For fabs spacex and Tesla are significant consumers of the output- vertical integration is preferred by lots of companies.
> I mean the short answer is that it doesn’t temperamentally suit him to do so.
Correct, and I think having a CEO who's totally unaccountable to a board and is temperamentally unable to focus on his core business is not good.
SpaceX and Tesla are nowhere near large enough consumers to justify vertically integrating a fab. The demand is predicated on Optimus producing millions of units per year.
This circularity will probably end up being seen in retrospect as so transparently absurd that it'll beggar belief that anyone accepted it.
Terafab profitability is really tough to get right, I'm open to ideas -- there's operating profit, operating profit net of amortization / depreciation, there's like six phases of the terafab specced out running through 203X (and as you say, slippage is inevitable there). I'd want to be considered a winner if the value of the EUV machines is up significantly, e.g. it was a balance sheet winner.
You've got to take Musk's forecasts with a massive grain (sack?) of salt. Pre-launch he was estimating making 250,000 Cybertrucks per year by 2025. So far, total sales over 3 years have actually been 60-70K.
I'd bet that in 5 years time Optimus total sales are in the 10's of thousands at best, with few to no large scale industrial users (other than self-use by Musk). Customers will be similar to what AgiBot and Unitree currently have - entertainment, hospitality, academic, small scale trials, plus a few Tesla fanboy sales that don't do anything actually productive.
On the cybertruck, I was going to suggest that's a function of demand, but actually I see there are substantial wait times for these. That might mean they cut factory capacity though, I haven't tracked it super carefully.
If Optimus is in use at factories in 2029 then it will absolutely be able to produce more than a thousand robots a month by the end of 2031, just using any sort of geometric production curve as a swag.
As to users, I think if it can provide home services, they're going to sell a ton of them.
As to Elon's forecasts, yes, double them at least. And, also, he's mostly right historically, if delayed substantially. I think there's near infinite demand for $1/hr robot labor in the US.
It’s so interesting. On the one hand you have our culture’s deserved rancor at late stage capitalist CEOs juicing the profits and squeezing every last drop of profit out of their customers. Tired derivative products and enshittified service models are exhausting us. Our institutions seem to be creatively bankrupt, and every single product and utterance they make has been through 500 different committees to ensure maximum profitability and minimum soul. We all fucking hate what almost corporate America has come to stand for.
And yet… when it’s Elon Musk’s company, we are suddenly all numbers and figures, and we waggle our fingers disapprovingly that he doesn’t have board approval. Yes, with Musk you have a childish huckster asshole. And increasingly it does appear he’s gone off the deep end, buying into his own self mythology. But the one redeeming quality he still demonstrates is refusal to settle for doing the same thing over and over. He continually and unreasonably pushes the envelope.
Tesla has definitively become a successful company. They defied all the naysayers in reaching that success, and it has turned into a global behemoth. They have oodles of money. The fact that Musk tirelessly insists on trying new moonshots that seem ridiculous is a GOOD thing - it’s the thing we _lack_. Inventive, audacious risk-taking is exactly what our stupid extractive scam economy lacks. Compared to you and me, yes, Musk is a cringey edgelord with the energy of a weird kid at the sleepover who tells bragging lies. But compared to his cohort who run large businesses, he is a goal oriented doer who attracts great talent and then relentlessly pushes them to achieve the impossible. And they actually succeed some non-zero amount of the time. It doesn’t matter that not every idea is a winner. That’s not how moonshots work. The point is that _some_ moonshot ideas do work though and actually move us up the tech tree. And the probable reality is that the delusional unrealistic asshole _helps_ push you there, precisely because they cannot be reasoned with. No, you cannot build moonshot companies without grandiosity, because the people working on it need to have a mission that is bigger than dollars and cents or common sense. If the only thing pushing you is good financials, you will quit as soon as “projections” start to look bad. In the area of moonshots, projections are just the opinions of uninformed non-believers who will ALWAYS mount until the world convinces you to stop working on the “stupid” thing.
All that said, the flying car is the dumbest thing I’ve seen him push and I don’t find it remotely exciting. Self-driving and robots are way cooler.
I don't really disagree with much here. I think Musk is a despicable piece of shit for reasons totally outside of his management track record.
However, the root of this conversation is not about Musk at all but about whether Tesla's core business (automotive) is in trouble. The answer to that, as Musk is demonstrably convinced as well as I am, is "yes."
Then we get into Musk's own personal idiosyncrasies and to what degree they're a liability or an asset to pull Tesla away from doom. For the reasons you mention, they're an asset. Not many leaders are willing to throw the wheel like he is (Zuck being another example). For other reasons, his idiosyncrasies are a liability. Like the fact that he personally was single-handedly responsible for destroying huge amounts of Tesla's brand power and that his personal project (the Cybertruck) inflicted such gargantuan direct and indirect cost on Tesla at such a critical moment.
Again, the question is whether Tesla's core business is in trouble. The answer is yes. Nobody remarks that it's such an incredible thing that a CEO of an unambiguously crushing-it company is willing to yolo the company into new directions.
Yeah - I guess not much I disagree with here. But, some food for thought: people on this website have been telling me Uber is going to fail for nigh on 10 years now. Constantly, breathlessly, they've explained to me it's NOT because they feel any one way about the misogynistic leadership elements at that company. No, they predict it will fail purely based on a dispassionate reading of the financials and fundamentals. It will definitely fail _because_ it's a bad company. Any day now, they assure me.
Point being: phrases like 'rationalization' and 'wishful thinking' exist for a reason. You should be extremely skeptical of how fairly you are assessing the prospects of a company run by people you hate. You are almost definitely interpreting the outlook uncharitably.
> our culture’s deserved rancor at late stage capitalist
Late stage capitalism was coined over a century ago. It still hasn't happened.
What does happen, however, is a company will grow until it strangles itself with committees and bureaucracy. And then gets replaced with a younger company. A good book on this is "The Innovator's Dilemma".
If anything his companies are doing well despite of him, not because of him. There's level-headed and competent leadership in them resisting his bullshit.
The counterpoint of course is that he has demonstrated a consistent ability to build and run successful businesses, many of which are extremely high risk
The only new western car company in decades to not fail, based on a technology nobody believed in at the time, least of all other car companies
SpaceX being a private space company, which was completely unheard of. No experts though they could succeed
Smaller earlier businesses like Zip2, paypal, etc
These are gigantic high risk companies with high failure probabilities, the chance they they all happened to be successful despite Musk and not at least in part because of him is stretching reasonable doubt for me.
I'm absolutely no fan of his politics or personality, but the man can run a business that nobody else even dares try (at least until he demonstrates viability and the market starts to generate competitors)
> The counterpoint of course is that he has demonstrated a consistent ability to build and run successful businesses, many of which are extremely high risk
Three sucesses.
Not so successful: Solar City, Neuralink, TBC, X.
(X exists, but before it was valued-and-bought by one of Musk's other companies, it appeared to have lost about half its value).
After 10 years, it is not particularly good to have to caveat that with "once it passes the FDA, assuming it's real".
The main problem isn't even the tech, though the tech isn't as good as it seems either.
Their main advance hasn't been the BCI: other teams also do this, main limiting factor is product lifetime rather than electrode count (though yes their count is higher).
The surgical robot is a much more interesting thing than the BCI. But even with the robot, "interesting" rather than "amazing" because (1) what else can that robot do? (2) what does it cost to use? (3) was it reliable enough for humans, or the reason for all the animal welfare issues? (4) some other BCI teams working on delivery via blood vessel and less invasive is better.
Main thing though for any sucessful medical tech is cost. Even with FDA (or non-US equivalents, somehow I don't even know the name of my local equivalent), medical interventions are famously expensive enough in the US to bankrupt and insurance in the US is famously reluctant to pay the full costs.
And up a level, the problem with Neuralink is their output suggests a lack of care to detail, of emphasis on speed over quality, and of having inherited Musk's personal attitude problem when faced with criticism.
TBC is The Boring Company, Zip2 was one of the three sucesses (and not IMO the x.com that became part of PayPal, that looks to me like pure luck for correct timing in the dot.com boom before the bust).
Hyperloop was a tech not a company and if I listed every tech he's done wrong I'd be rightly accused of misunderstanding how R&D investments are necessarily gambles.
paypal succeeded _despite_ musk and required a boardroom coup to remove him from leadership because he was about to tank the company. if he had is way nobody would remember paypal besides a couple people with merch for a long defunct company.
I'd still restrict this to an ability to pursue ideas that others find too risky with relentless persistence and a laser focus. He is able to get something off the ground and make others believe in it, but not good at actually running a stable business. And that's fine, mind you, I absolutely respect what has built. But the person who is responsible for turning SpaceX from an incredibly risky bet into a profit machine is Gwynne Shotwell, not Elon Musk.
Of course it’s still early days for all these companies and they’re not exactly fiscally solid, but if Tesla had not made all those cars, someone else would have.
I ran into this trying to find a good book on modern python. There are tons of releases on very specific python versions, and when I dug into it they were all written by AI and reviewed poorly. It's honestly just getting hard to find good technical books in certain areas because there's so much nonsense going on in the space.
People forget the fact there may not be enough capital in the world to invest in AI, and so while some companies will go under the market could grow for literal decades. I don't know the future, but this is not a guaranteed bubble.
As the article states there's no reason to think Starlink temporarily slowed down their internet service deliberately. Elon has been a strong supporter of the Ukrainian side, donating hundreds of millions to them and risking retaliation from Russia against their satellites. I understand why they're upset, people can die when their internet is disrupted, but the article seems especially hostile for what is likely a bugged update or Russian interference on their military-grade service.
The reason to think that is because Elon previously personally intervened and neutered Ukraine's Starlink after a Russian government official told him Russia was going to nuke someone.
Elon is so fucking stupid he doesn't know how nuclear deterrence works and he actually thought Starlink was the deciding factor.
Shame on Elon for saving the world from nuclear holocaust.
For what its worth though he denied extending their network into Russian territory due to a threat of a nuclear attack. Not only did Ukraine know about it, but it was a stop on internet service. The disruption described in the article sounds more like the quality of service issues the way they continued to operate on reduced speeds and flaky service.
I know this is completely lost on Elon and the likes but the attack was not on "Russian territory" it was on territory most of the world does not recognize as Russia.
It's possible, I don't know for sure, but just the way the article described it this sounds like a quality of service issue, which would suggest technical difficulties. Starlink has a button to turn off service, but they probably don't have a button to indeterminately flicker internet speeds down and up with randomly high latency.
What would be nice too is if it lowers the bar for entry, then people closer to the real problem can work on it. There are a lot of mediocre or overpriced products due to the fact a team of software engineers don't understand the problem well
100%. A lot of the time the people who understand a problem the most are the people who work in that field/industry every day, not the tech folks that often do the work/startups that try to 'revolutionise' said industry.
The vibe-coded product the actual users put out will probably be nowhere near as good as a professionally built one tech wise, but will actually be built to solve relevant pain points.
I wonder if we’ll see a reduced feature set in Europe as well, especially when it’s first released. Honestly it’s incredible it’s taken this long for it to come out
Incredible, and here I was thinking the letter was going to be in regards to the last minute extension of the election by Lula's former attorney, or the irregularities in the second round. Facebook is a culturally left-wing company and their argument seems to lack substance
reply