A.

@NisforNora Go to the “history” tab on the sidebar.

How can I see my likes my twitter please 😩

Maybe Codex is better today. That is not the point. AI is not ultimately a demo contest. It is an infrastructure war. Models get commoditized. Compute, power, cooling, deployment speed, and eventually orbital capacity do not. The player that can build AI infrastructure fastest on Earth and off Earth will have the deepest moat. If that is SpaceXAI, then betting against Grok in the long run is the shallow take.

there is no musk vs altman argument when it comes to product quality. there is no way grok will be better than codex in the near future.

@Ssaasquatch 💯

@Brainmetry Fair enough. I guess only time will tell. Good convo 🤝

Space X trailing PE ratio is 1000x which essentially means earnings need to compound 30% for around 20 years before it grows into its valuation. How does Elon always get away with this? Lmao.

Fair point: a 1,000x trailing P/E absolutely means a lot of future success is already embedded in the price. And fair again: megawatts are not the same thing as returns on capital. Where I think your framework is too narrow is that it treats every new layer as if it must already look like mature AWS-style economics before the market can assign it any value. Anthropic did not go to some random GPU host. SpacexAI leased Colossus 1 capacity to Anthropic only after moving its own training to Colossus 2, and reporting described Colossus 1 around 500 MW with the broader buildout being pushed toward roughly 2 GW. That does not prove AWS economics. You are right about that. But it does prove something important: when frontier labs need serious compute, they go where real infrastructure exists. And right now SpaceXAI is clearly in that conversation, even while other hyperscalers and AI labs race toward their own gigawatt-scale campuses. That is why I think you are underweighting the bottleneck. Frontier models are becoming more repeatable over time. Frontier infrastructure is not. A lot more teams can train strong models than can secure power, land, cooling, chips, and deployment speed at this scale. Same thing on the space side: Starbase has approval for more annual launches, the Starbase footprint has been cleared to expand, and Florida is getting major Starship infrastructure too. So Starship is not just ambition anymore. The physical network for much higher cadence is being built right now. I agree Starlink is the cleanest earnings bridge today. I also think the market is right to price upside from Starship dramatically improving deployment economics over time, including for a much larger next-generation Starlink network. And if orbital AI data centers happen, SpaceX is the only player with an obvious path to deploy multi-gigawatt compute off Earth cheaply enough for the idea to matter at all. That is not base-case cash flow today, but it is rational optionality for a company building both the launch system and the compute stack. I would not even need X to make the core SpaceX case. If X’s monetization improves, that is extra upside, not the main thesis. So yes, great companies can still be overvalued. You are right about that. But the opposite error is demanding mature cash-flow proof from businesses that are still building scarce industrial bottlenecks that everyone else may depend on later. SpaceX may be expensive. I just think your framework understates how valuable it is to own the bottleneck itself.

Okay I’ll indulge you. You’ve actually not given me a sound rebuttal but only listed out a few exciting potential catalysts that still doesn’t justify its valuation. Your mature-company metric critique is funny. The point of referencing trailing P/E is not to say SpaceX should trade at 20x earnings but to show how much future success is already embedded in the current valuation. When a business trades at 1,000x trailing earnings, every bullish argument needs to spelll out the level of future earnings that makes its current price rational, and how long does it take to get there? Until that math is shown, SpaceX is overvalued. Yes trailing PE may be imperfect for an early stage company and that’s exactly why I juxtaposed that with future earnings growth to put things in perspective. At 1000x PE, the burden of proof becomes absurdly high cos you’re not only pricing future potential but flawless execution by Elon. “SpaceX is more than a rocket company” is also not a good argument. Even if we call it an infrastructure platform then value the actual cash flows of each layer, Starlink has to justify its economics (which it’s already doing to be fair), Launch has to justify its margins. Starship also has to move from mere ambition to proper commercial scale. Defense has to become material. The point is that XAI infra has to have real monetizable capacity and not just another capex-heavy adjacency with growth assumptions baked in. Can they grow 30% yoy? Only time will tell. Also, hyperscaler-scale power consumption is not the same thing as hyperscaler-scale economics. A 500MW or 2GW buildout only tells me the ambition is enormous but says nothing about the potential ROIC which is the koko. Capex intensity can create value, but can also destroy value in the same manner if the market is already capitalizing the upside before the real actual returns show up. Your Anthropic/XAI point is even funny cos Leasing spare capacity is great, but does not magically turn SpaceX into AWS. One customer using excess infrastructure does not prove durable pricing power, high margins, or a defensible cloud business. It proves there is demand for compute. We already knew that. Heck, every idiot knows this. Finally, nothing against SpaceX btw and it may be one of the greatest companies ever built but it could still be overvalued at the current price. Those two truths can exist. The market does not pay you for identifying greatness. It pays you for buying greatness at a price where the future has not already been fully capitalized.

@_masterinvestor Good rule for most IPOs. SpaceX is the exception people will realize too late. It gets underestimated because people still value it like a space company instead of what it actually is: a compounding industrial platform with no real peer.

FOR EVERY IPO ON EARTH WAIT 6 MONTHS

Free profit doesn’t exist when you’re paying a middleman 25% of your revenue just to list your cars on their app. Tesla isn’t a wild card, they are a vertically integrated machine. They own the hardware, the FSD software, the Supercharger network, and a massively active app with millions of daily users. Uber knows exactly what is coming. That’s why they are signing deals with every AV company they can find. They are terrified of being cut out.

@Brainmetry I’d say existing human travel patterns have created the peaks and valleys. Robotaxis def love graveyard shifts more than humans - Uber will use them. So should every manufacture attempt to cut out distribution? Why not, it’s free profit right? Tesla remains a wild card.

The Uber is 'just an app' crowd is out in force this weekend. Let's be intellectually honest for a moment: it's a network that runs over an app and that network is fractured on both the demand and supply side. $UBER

Humans are the reason Uber has a variable supply problem. A human driver won’t commute just to work a one-hour surge. A robotaxi will. During the valley, AVs don’t sit idle burning human time, they charge when electricity is cheapest. And questioning if a hardware manufacturer can run a consumer service ignores reality. The Tesla app + Supercharger network is already one of the tightest hardware-to-consumer software loops in the world.

@Brainmetry Low cost cars push back against the variable supply problem but don’t eliminate it. Uber rents cars very cheaply from owners & still needs massive flexes in supply given very intense peak to valley swings. Hardware mfrs running a high functioning consumer service…maybe.

Fair point on the transition period. Uber is essentially the training wheels for the AV industry. But the variable supply problem assumes vehicles remain expensive. If you bring the capital cost of a robotaxi down to $25k - $30k through scaled manufacturing, idle time stops being a financial death sentence. Uber provides liquidity today, but they are training their own replacements. As soon as the hardware makers hit critical mass, they will cut out the middleman.

@Brainmetry Permanent moat is not in my vocabulary. Don’t disagree with your framing and AV economics have a variable supply problem that is not an easy solve. The most liquid provider would seem to have an important seat at the table especially over a long transition period.

@Jason @ElizabethHolmes @SBF_FTX @stablekwon SBF. Biggest story, biggest fallout, biggest cultural impact. If you are making your first documentary, start with the one people still cannot look away from.

Who should I make my first documentary about? @ElizabethHolmes @SBF_FTX @stablekwon or other?

If Drake leaves UMG, his back catalog stays with the label. Every new album he drops, on any label, drives revenue back to music UMG controls. The smarter play isn’t to leave but to negotiate a renewal that includes a path to owning his masters.

@yatharthmann The difference between an orbital and a suborbital rocket is like heaven and earth.

To be fair, Jeff Bezos has also been landing (sub-orbital) rockets since 2015.

Ridesharing is a commodity. People don’t care about the Uber app; they care about getting from point A to B for the lowest price. If Tesla drops the cost per mile by 50% with mass-produced Cybercabs, it does not matter if Uber has a 10-year head start on the app store. App monopolies disappear overnight when the underlying hardware changes.

@Brainmetry It will take ten years for this deployment to happen, and whoever buys Uber will go right to the front of that race.

Uber is going to be bought by Google/Waymo, Amazon or Tesla/SpaceX in the next year. For a “buy it now” price of $250b, one of those three companies gets a $12b a year free cash flow machine with $70b in revenue — and hundreds of millions of global customers This is the most obvious M&A deal since Instagram, Android and YouTube transformed Meta and Google Discuss

Brilliant idea. Let’s go ahead and ban space exploration and satellite internet. Actually, let’s just pause all human progress so the “experts” can finally relax. We can all sit in the dark, do absolutely nothing, and slowly revert to the Stone Age. I’m sure the media would love that. It is crazy how the people who build absolutely nothing always have the most instructions on how to run the future.

Satellite launches are filling up the atmosphere with pollution, study warns trib.al/FIucxfw

@JCChristopher It’s disheartening how many people have lost their sense of reason and simply resort to repeating buzzwords and talking points that make them feel better.

Honestly, I just don’t give a fuck. He came to the wrong Tesla channel. But make sure you like and subscribe! I do care about that!

Dude, you’re a complete moron. They presented you with facts and figures, including the Netherlands’ rigorous testing on their road for a year before approval. He even encouraged you to test it yourself, but hatred and stupidity prevents you from reasoning rationally. You guys are insane.

@minzirillo @ElectricRaph @AnnickDeRidder @KRoelandschap The only facts you posted is Teslas own misleading and debunked data. RDW is the first notable third party assessment, though it is still intransparent.

What a wonderful development for FSD (Supervised) in Belgium 🇧🇪 Thanks to @AnnickDeRidder for bringing this to the agenda in Flemish Parliament. She is well-equipped, competent and knowledgeable as became very apparent in her speaking to the Parliament. Very curious how this will move forward! God speed Belgium! 🇧🇪 x.com/dm_jns/status/…

Weeeelllll, I guess @Starlink Mobile is doing something right! It’s David and Goliath (X3) all over again — I’m bettin’ on David :)