SpaceX says it has agreement to acquire Cursor for $60B

2 months ago (twitter.com)

https://www.reuters.com/technology/spacex-says-it-has-option...

https://www.nytimes.com/2026/04/21/business/spacex-cursor-de... (https://archive.ph/c2Tac)

https://www.bloomberg.com/news/articles/2026-04-21/spacex-sa...

So SpaceX bought a $60B Option on Cursor, plus a bunch of services, for $10B.

If strike date comes and Cursor is in fact worth less than $60B... they can move to acquire it for that price. Or just let it "expire". And if it's worth more, they get a savage good deal. If the services were worth $8B anyway, it's hard to lose.

It seems less crazy to me through this lens. A straight acquisition, today, at $60B would in fact be crazy.

  • What's crazy is that a company that sells an IDE (that's not even a particularly good one compared to competitors like JetBrains) integrating some AI plugins could be worth more than $60B...

    • In terms of IDE yeah it is not that great.

      I do have Copilot in VSCode and Cursor.

      I thought both should be equal in solving problems - turns out Cursor with the same model selected somehow was able to solve tasks that Copilot would get stuck or run in loops.

      They have some tricks on managing file access that others don’t.

      29 replies →

    • They are now a Codex clone and without the subscription pricing. You have to spend thousands to get what you get from a $200 Codex subscription. How do they compete with this except from users who haven't caught on yet, or businesses that are unbothered to spend thousands a month per dev and wouldn't consider just subscribing to 1-3 $200 subscriptions instead?

      And their price is so high because it's markup on API rates. API rates, even without markup, are just insanely irresponsible for anyone to be spending on full-time daily usage.

      27 replies →

    • the IDE has little value

      What they want is the massive user base, the data (Cursor has a lot of high quality coding data for training), the teams expertise in coding models and agents, and the Composer models

      60 billion is a large number but these frontier labs are burning billions a month in compute alone, and SpaceX is IPOing soon so they'll have a lot of cash to spend

      7 replies →

    • > What's crazy is that a company that sells an IDE (that's not even a particularly good one compared to competitors like JetBrains) integrating some AI plugins could be worth more than $60B...

      yes. plus $2b ARR, 1m DAU

    • > that's not even a particularly good one compared to competitors like JetBrains

      Massive understatement calling it "a not particularly good plugin". If it were that simple there wouldn't be a need to even do this.

  • Paying $10B for the option is also crazy though. Paying $10B for the thing outright and not just an option would be absurdly high.

    • Is this cash or compute? Elon has one of the world's biggest compute clusters spun up, and little inference demand to speak of.

      Trading billions worth of idle compute, in exchange for a high-strike call option on the #3 player in the most-promising-vertical for AI, plus (presmuably) some access to their data, starts to sound like not a bad trade. Especially if you're pre-committed to betting your entire rocket company on winning in AI, and you're currently in sixth or seventh place.

      33 replies →

    • Problem is basically, that if the option works out (Cursor truly has the talent to train a frontier model on SpaceX's infrastructure, and were simply lacking the infra before) the fair price would be way way more than $60B.

      OpenAI tried to acquire Windsurf last year for $3B and couldn't.

      4 replies →

    • Not only is it almost certainly compute (“services”) it’s likely priced at Anthropic rack-rate, or at least what Cursor’s been paying Anthropic.

      The cluster’s already paid for, so likely in the $2B range for operating cash needs. Not more than $5.

      If I imagine bringing in Cursor’s team to build a frontier model, ideally combined with Grok, which has one of the few truly proprietary data feeds available to it, and with a much larger custom model Cursor can solidify a place, and I get to do a stock swap to buy it, this sounds like a bet worth making.

      Upshot - I bet there’s an MS/oAI deal on IP on the back of this; meanwhile the cluster goes brrr.

    • But they also get a whole bunch of AI Services from Cursor. Other comments have noted that xAI has fallen on bad times (idk one way or the other) so perhaps they were going to spend $5B on getting these services elsewhere, anyway.

      SpaceX spending $1B a month on various AI services seems ~plausible

      (EDIT - Or maybe it's an IP transfer, or maybe it's over a longer time horizon. Idk but SpaceX clearly expects value from 'our work together' even if they don't exercise.)

      7 replies →

  • Despite their impressive ARR, Cursor faces existential threat from not only BigLabs (Claude Code, Open AI Codex) but also BigTech (AWS Kiro, Google Antigravity, MS VSCode). I am sure the usual suspects would have lined up to purchase Cursor, and the deal from xAI was probably the best of the lot. Marks an end to a remarkable sprint for a 3yo company, and an admirable exit (considering the recent discombobulation of Windsurf's), just as investor money and/or hype is going belly up.

    • Having tried most (all?) of the commercially available + open source options, and even tangential competitors like CC, Conductor, Antimetal, etc. I haven't found anything that's close to the experience of Cursor. The harness they've built is incredible.

      I'd even go so far as to say that any competitors that are direct (windsurf, kiro, etc.) aren't even in the same universe. Cursor is just so much better, faster, has better features (plan and debug mode), and squeezes much better results/code out of the same models. They absolutely have some secrete sauce that the other options just don't have.

      2 replies →

    • It has shown surprising stickiness. Occupying some middle ground between full adoption and still ~in the code.

      I am starting to see some potential in moving back away from pure terminal, a mixed modality with AI. But it is not in the direction of IDE in any traditional sense.

      1 reply →

    • Do you really think anyone is using AWS Kiro or Google Antigravity? They are not real competitors in the slightest.

  • This valuation is absurd. Perhaps a year ago- sure, but there have been so many iterations of this “kind of editor” since then, not to mention countless alternatives.

    So for me it’s more of a data deal - Elon buying himself some insight into codebases and real dev usage patterns? Oh finally someone to use his dirty data centres

  • If you pay 10B for options at 60B and the strike is 8B you ... just lost 10B. Thats it.

    Add emotional hedges if needed but they are just emotional not financial.

    Your argument is based on an assumption that cursor cannot lose value. Even if the market says it has.

    No free lunch: an option is a bet for both sides. Zero sum.

  • 3 things bug me Now why would cursor agree to that unless the offer was better than what their market valuation + acquisition premium < 60

    This was a similar play for twitter by the same person

    While an innovator at the time, today there are a lot of LLM coding solution, sold by model providers, model aggregators even open source ones , it’s not obvious what is being bought that isn’t a feature of vs code or one of the LLM agents ( as the dismissive saying goes )

  • I used to be a hardcore Cursor user until I spent more time with the alternatives and cursor has gotten rough on my machine lately: spins up the MBP fans in minutes and drains the battery in an hour or two on the go. The harness is great, but eventually someone will build something equally good that is not vibe coded to death.

  • Thinking of it as an option makes it much more rational.

    Downside is capped (cost of services + deal structure), upside is asymmetric if Cursor outperforms.

    That said, these deals always hinge on whether the “$8B in services” is real economic value or just internal accounting. If it’s the latter, the risk profile looks very different.

knee-jerk is that it's weird, but makes sense:

* X will have a total of ~2GW of GPU sometime this year largely not doing much outside of 'grok is this true'

* despite no longer being in vogue with consumer devs Cursor still has a lot of developer data that can assist in building a model

* Cursor have decent enterprise relationships (while for xAI it is ~zero) and that's where the real revenue for llms + agents is

* Cursor are paying retail for tokens and competing against the frontier model co's who are also their suppliers. Not sustainable (hence their in-house composer model).

* Cursor the product covers the gamut from lovable-style prompt-to-app, an IDE, cli and bugbot

* X are using "x bucks" to pay for a potential later acquisition which are arguably overvalued based on the space x IPO hype

Option there to give X a window to make it work, otherwise walk away with a $10B breakup fee for access to it's data

  • > largely not doing much outside of 'grok is this true'

    Hey now, don't forget about it's super important other use, taking innocent photos of people and regenerating them in less clothing and compromising positions.

    I'm sad that I even know that.

    • They changed that recently, you need to be paying €10/mo for that now. The free plan and/or access for the basic Twitter plan are gone.

      6 replies →

    • Photoshop has been able to do that for 25 years. Do people realize that AI doesn’t magically know what their real bodies look like? AI is just pasting the averaged body parts from every porn image on top of what you were wearing. I know people love to be offended, but it’s weird to me that they’ve made up a right to not have someone privately mess with an image that has your face in it.

      Maybe if this tech were completely secret and this was 1997, so a video of a naked Bill Clinton high-fiving Saddam Hussein in a hot tub was likely to shock the world, then it would be a big deal. But everyone knows all images (and especially surprising ones) are likely to be AI, I’m asking sincerely, does it really matter if people make fake photos for wanking purposes?

  • I wouldn't be surprised if those enterprise relationships evaporate after this acquisition. There's a reason why xAI has zero enterprise customers.

    • For Enterprises it's way easier to delist Cursor from the list of used tools than to have a relation with someone known publicly for neofascist aspirations.

      xAI is not, and was not that bad, it's just everybody ignores it for anything serious due to obvious reasons.

      2 replies →

  • You've literally got tools like opencode that are MIT licensed. Most of those points X could do on their own or are things that make this attractive for cursor not X.

    e.g. Need developer data? Use some of that spare GPU compute, hand out free top end model coding access for a bit and you'll very rapidly have developer data

    >decent enterprise relationships

    I guess. 60B worth of "relationships" though?

  • Yes I think you're right. Reinforcement learning is extremely compute heavy, which cursor doesn't have. And X.ai doesn't have the coding agent data anthropic/OpenAI has, but does have the compute.

    However, one thing in AI is that while the usage goes up extremely quickly, it tends to go down just as fast. I know a lot of companies that are in the process of switching from Cursor to Claude Code, so in 6-12 months I'm not entirely sure of the data quality/quantity.

    Also I think it is telling that they are calling them SpaceX not X. The X brand is absolutely toxic, especially in enterprise.

    • > Also I think it is telling that they are calling them SpaceX not X. The X brand is absolutely toxic, especially in enterprise.

      it might not help all that much once it turns into "grok" harness or otherwise associated with elon

  • I think you're right. Other providers can offer coding subscriptions that use in-house models, and this sets the stage for a Grok coding plan that's built in to Cursor.

    $60 billion seems expensive, but it gives them a much better chance at competing in the market than if they started their own harness from scratch.

    • Absolutely no enterprise - I work in enterprise cloud consulting - absolutely no company would trust Grok with their IP compared to Anthropic or OpenAI with Musk’s reputation on how he runs his businesses.

      Anthropic just tolerates the money losing developers who pay $20/$200 for subscriptions.

      3 replies →

  • You forgot to consider whether all this is worth $60B.

    • > forgot to consider whether all this is worth $60B

      I see two possibilities:

      (1) SpaceX is paying with stock; and

      (2) the $60bn pay-out is (a) conditional or (b) never going to be exercised—it was a stalking horse for negotiating the $10bn terms, which gives SpaceX everything it actually wants.

      1 reply →

    • $1B to $2B ARR in a few months with projection of $6B ARR by years end. If xAi wants to have it's own tools just like OpenAI and Anthropic, then it's not an unusual move.

      3 replies →

  • I think it also represents a bet that in some sense Cursor's model capabilities are resource limited rather than talent limited. If that's true, $60B will end up being a bargain. If not true, well it's an expensive lesson but that's the nature of things.

  • Forgot that claude is burning good will from it's own capacity constraints, leading to periods of 'dumbness'. It's a catalyst to cause me and others to switch back to cursor if they can get their act together

  • > despite no longer being in vogue with consumer devs Cursor still has a lot of developer data that can assist in building a model

    care to share more about this?

  • > Cursor still has a lot of developer data that can assist in building a model

    Their composer model is seriously good. I’ve been eyeing a cursor sub just to use it in OpenCode. They have a nice moat here.

    > Cursor have decent enterprise relationships (while for xAI it is ~zero)

    That has a reason. Those enterprise relationships are almost certainly going to sour at least a bit, if not for Musk‘s toxic image then for his erratic behavior.

    • Just to point it out, Cursor has not made any good models themselves. Composer 2 is Kimi K2.5, and they tried to pass it as their own until people noticed that the api specified it as Kimi.

      3 replies →

    • Yeah, Composer 2 is legitimately so impressive. It is my daily driver right now both on professional and personal projects. I only find myself reaching for 5.3 Codex/GPT 5.4 when exploring a lot of technical documentation or code and for Sonnet/Opus when working on UI. Everything else is Composer.

    • Even if it wasn't for Musk, are these relationships really worth so much? There is a certain value in being on the approved vendor list, but it seems to me that there really isn't a lot of vendor lock-in. I think most people could switch to opencode, claude code or codex pretty easily. Maybe these relationships would be worth a lot if companies signed long-term contracts, but I doubt many did.

    • Why not buy OpenCode subscription? You'll have access to better models. glm5.1 and qwen3.6 have been really good for me

Why would Elon do this if he knows full well the names X-Code and Codex are already taken?

I guess the hope is that combining two sub-par coding models (xAI's grok + cursor's composer) and combining the data they have access to, they can build something that can compete with OpenAI / Anthropic in the coding space...

I guess I kinda see it... it makes sense from both points of view (xAI needs data + places to run their models, cursor needs to not be reliant on Anthropic/OpenAI).

I think I don't see it working out... I just don't see an Elon company sustaining a culture that leads to a high-quality AI lab, even with the data + compute.

Guess I'll be looking for a replacement for Cursor now...

Anyone have recommendations? I like the plan/agent mode and the fact that it's an IDE, so I can use it in the traditional way as well as by yapping with a bunch of agents. Also the Cursor rules I've curated and they do their job well.

  • Yep this was the moment to finally remember to cancel my cursor subscription. I find it unconscionable to do business with someone who would do business with Elon.

  • Zed - https://zed.dev/

    Integrates a lot of agents (I use it with OpenRouter and directly with Pi) natively, is fast (you don't realise how laggy VSCode and its forks are).

    Biggest disadvantage: lack of extensions. Lots of quality of life missing (e.g. gitignore integration to add/append gitignore files for different languages).

  • any IDE you like and Claude code - i have no idea why you'd want to use something like Cursor, it's time came and went.

    • Afaik Claude has 5 hour rolling window rate limits while Cursor has a monthly window on the $20 plan.

      The 5 hour window sounds annoying for hobbyists who only use it time to time when they want to dive into some personal project.

    • Not to mention the only company that would have any legitimate interest in acquiring Cursor would be Microsoft since they could just merge VSCode and Cursor into one product at very little cost.

  • Ironically I cancelled cursor a few days ago. I went back to good old VSCode and just use the Claude code and codex extensions.

  • Consider Nimbalyst, its a free visual workspace for Claude Code and Codex that has visual editing of markdown, mockups, diagrams, code with your agents with WYSYWIG diffs as well as task management and kanban session management tied into your agents. Its got a files/plan/editing mode and an agent/sessions mode.

  • Until yesterday I would have recommended VSCode + Copilot. They had the best pricing of any option. However the pricing was unsustainable and is therefor finished.

    • How's the ai autocomplete? It was unusable in November when I tried it last and went back to Cursor. Slow and when it finally did something it was just not good. Cursor is super fast and actually gives useful results. I just don't want to give my money to Elon, so I might cancel anyways.

      4 replies →

  • In the same boat and landed on Kilo code.

    Similar to Cursor, open source, based on open code, compatible with claude code configuration and works with all providers.

    Seems quite future proof to me and you can toggle between providers seemlesly

  • (disclaimer: I work at ampcode)

    Give the oracle at amp a go :) Our TUI is really nice as well. Get in touch for some credits.

I'm sworn off from Musk-related products, and this will prob make cursor worse (switch to X's LLM for instance). So, any suggestions for switching? Codex; Claude Code? (I like my IDE and I like the freedom to choose a model, which is why I stuck with Cursor even when it felt more expensive)

  • Zed is snappy as an IDE, and ghostty for your CLI. I've done like 99% of my work in the past month just in ghostty + CC.

  • Everyone talks badly about Cursor and it is kinda a piece of junk, but no, there's nothing that has the features of: being able to see agent diffs in an editor, seeing diffs inline in chat, be able to click them to jump to the code, and being able to click old chat messages to edit/fork them.

    Those are basically my only requirements, and it feels like I've tried everything and they're all everything only has 1 of those features. Zed is the closest, it technically has those features, they're just buggy and have provider specific quirks.

    So I'm stuck on Cursor until Anthropic invents IDE technology, or at least VS Code wrapper technology.

    • Jetbrains IDEs have AI support with all the things you've described, and in a more polished experience that requires significantly less maintenance and tuning. It does that while affording an actual IDE experience that works well for supported languages/projects out of the box, without the need to constantly tune plugins and experience jank misaligned UX that seems to be the norm for VSCode and derivatives.

      No association with Jetbrains, and despite having a license, don't even use their AI support much myself (mostly using CC, with IDE integration for diff viewing). But if you haven't tried it recently, probably worth a revisit if you're open to Jetbrains products.

      1 reply →

  • I really doubt they'll swap in Grok. Grok seems pretty dead. Probably more likely they'll reuse the hardware for composer.

    If value is a concern, Codex. It's pretty hard to beat those subsidies. If you really want model freedom, Copilot is surprisingly decent value and as of right now let's you use your sub in other harnesses like OpenCode.

    • Codex is not a replacement for an IDE. Yes I still need an IDE.

      When coding agents work they're great. When they don't I still need the IDE. They usually don't work that great when I'm working on something novel or brownfield. Which happens quite regularly.

      But I definitely still want ai autocomplete. I'm not a Vim user. Coding isn't about typing for me, it's about solving problems. So a tool that does lots of the typing for me is a godsend.

      So do I go for VS Code + Copilot? Because it was bad when I tried it again for a few days in November. Slow to respond and gave poor results. Cursor is snappy and gives useful results most of the time.

      1 reply →

  • I switched to Windsurf recently and it's been pretty good for providing a Cursor-like experience; pricing is pretty similar.

  • Same I'll be switching off Cursor today to either claude code or kiro. Luckily my company lets us choose which agentic software we want to use. I won't touch anything Musk is related to, he is toxic and anything he touches turns toxic and supports him.

  • Kilo code; vscode extension as well but open source and based on OpenCode.

I was required to use Cursor for my job when I first started, but once I figured out how to use the command line version of Codex, I kind of stopped seeing the point. It just kind of seemed like a bloated, overpriced wrapper around what I could do with the included ChatGPT membership I already had for work.

Maybe I was missing something, but I do not understand how it is worth sixty billion dollars.

  • It's not. It's a glorified code editor with no moat. Those are (massive) bubble prices.

    • and I'm being completely neutral and objective in saying this: Elon Musk has been a horrible capital allocator but great at financial engineering. X is still struggling to win back advertisers (they will never come back) and still in the red. I have little reason to believe this is also another careful and shrewd financial decision.

      2 replies →

    • I mean, technically they also re-sell AI tokens. Unsure if that’s with a markup or a discount.

  • I do think the Codex harness is a bit better than others. Doesn't make a ton of difference with OpenAI models, but with Google and Anthropic models the difference is quite noticeable I think.

Anyone saying this is an aquahire has it backwards. SpaceX is acquiring Cursor’s customers, all those enterprises including NVIDIA itself. I believe Jenson Huang is on the record about the engineers using Cursor everyday.

As far as I know, xAI’s enterprise market share is non-existent. This is their way to get some much needed customers.

  • NVIDIA has 42,000 employees. Even still, when their deal with Cursor comes to an end do we really expect them to stay loyal? And further, sign on with xAI?

    When they could instead sign with the new hottest enterprise coding IDE (Claude, Codex, etc who are way more popular now). Maybe if it’s an acquihire, it’s the GTM/Sales that xAI is after?

    • They might. Elon will probably use his SpaceX/xAI spend (future SpaceX space datacenter dollars) as leverage. NVIDIA is so used to doing such deals by now, they’ll probably take it.

      2 replies →

    • at 42,000 employees and their own (infinite) compute on hand, there has to be at least one plucky junior internally who is suggesting using the open source equivalents, internal / open models and saving a big pile of money.

Lots of people in the comments talking about how this is about training data, but surely this is actually about hiring competent people after the mass exodus/firing at xAI?

  • Whoever thinks the talent pool is this limited that it requires offering Cursor of all places $60B is pattern-matching so hard they might as well be a quilt.

    • It's not about the talent pool at all.

      The AI bubble music stops when one of these former-darling companies has a complete crashout rather than a "successful exit". The investors keep investing because even largely-failed products get acquihire paydays

    • hi, im the quilt.

      Note that Meta paid ~16b for Alexandr Wang, and Google paid ~3b for the windsurf executive team. You are making a category error -- the talent pool isn't "ML researcher" it's "competent leader"

      2 replies →

  • They could offer $20 million dollar signing bonuses to every Cursor employee if they wanted to hire them away and it would be much cheaper.

    They’re buying the customers and the brand.

    • Buying the customers seems though, when it looks like they migrate to whomever offers the steepest subsidies.

  • The acquihire angle is probably part of it, but I'd note that Cursor's team is small — around 50 people — and the $60B valuation makes it expensive per head even by AI acquihire standards. You don't pay that multiple for talent alone.

    What you might pay for is market signal + model distribution. Grok needs a story for why enterprises should switch. "The model that powers the tool you already use every day" is a much easier enterprise sales pitch than "our LLM benchmarks slightly better." The $60B is at least partially buying the answer to the question: why should any company bet on xAI?

  • Are cursor developers “competent” in creating frontier models? Aren’t they just using other company’s models?

    • I think composer has currently by far the best price to performance ratio for coding (not counting subsidized subscription cost by OpenAI and Anthropic). It's based on Kimi K2, but I think it's fair to say, that their RL really sets it apart from the other open weight models.

Bloomberg reporting its an agreement to either acquire for $60B later this year or pay $10B to work together https://www.bloomberg.com/news/articles/2026-04-21/spacex-sa...

Cursor's statement on the deal (which does not mention the option at all): https://cursor.com/blog/spacex-model-training

  • It sort of implies the $10B is going to be paid with compute credits. So this could very well be xAI simply compensating Cursor by giving them $10 billion worth of tokens. (What’s a token worth in dollars these days?)

    • That’s indeed the trick. Spacex “invests” in Cursor, looks good on their balance sheet.

      And xAI now gets 10B of more revenue on their income statement.

      Perfect financial statement boosting for the IPO which in turn will pay back these costs.

      At least that’s the bet.

      4 replies →

I just want to make the observation that this whole SpaceX IPO is turning out entirely unlike the CDOs that led to the 2008 financial crisis. There's no mixing of AAA level assets with a bunch of subprime stuff and then getting someone to buy it all as AAA. Not at all similar. Completely different. Will turn out just fine this time.

  • The assets weren’t AAA, you’re mixing it up a bond concepts. The deal had bonds that were AAA. And if you’re talking about CDOs then the assets were bonds which were usually BBB or similarly cuspy bonds.

    You should learn about securitizations. It’s actually interesting. But people talk about it colloquially and so incorrectly that it’s mind dumbing.

    Here’s a simplified example of how you can take something and turn it into a safe investment:

    Suppose you have 10 loans and each has a 50% chance of default. Ignore coupon, and say they are $10 each. Expected value is $50

    If you were to put this in a deal and cut it up into tranches, say the first tranche gets the first $10, this would be your AAA bond because odds of getting paid out you $10 would be > 99.9%. The equity (bottom tranches) would pay a lot less. For instance the expected value of the bottom half would be considerably less than $50 that is being promised. So there’s upside since you’ll be paying cents on the dollar and even though in the median scenario you’re making nothing, you have to weight the expected values of each scenario to figure out how to price it.

    The problem w this model is that it only works if assets are relatively uncorrelated which wasn’t true (it was true in the past but ignored systematic risk and adverse selection in originations).

    What this has to do w musk or spacex I’m still not sure

    • Just to well actually your well actually...

      What you've described is how the base level mortgage-backed securities (MBSs) work. The tranches work because there actually exist mortgages that are at lower default risk (high home equity, well qualified borrowers, etc.), and the senior tranches are effective in capturing their underlying safety. What CDOs did was to take the lower, riskier tranches of MBSs from various sources and repackage them and divided them into tranches again. Then they got the ratings agencies to rate the top tranches of the CDOs as AAA as well. It's as if a teacher graded several classes and then took everyone that got a C or below from all the classes and then graded them on a curve again. And suddenly a lot of the C students became A students. It was outright financial insanity. Well, mixing a rocket/satellite company with a couple of also-ran AI outfits and the walking corpse of Twitter, and then calling the whole thing SpaceX and valued at $1.75T is a similarly level of financial insanity to me.

      1 reply →

    • > this would be your AAA bond because odds of not getting paid out you $10 would be > 99.9%

      I think you meant "the chances of getting paid", not of not getting paid.

      1 reply →

    • > Suppose you have 10 loans and each has a 50% chance of default. Ignore coupon, and say they are $10 each. Expected value is $50

      And that naive statistical reasoning is where it goes terribly wrong. You have to consider the causal process that generates that distribution!

      The type of people who would default on a coinflip are extremely sensitive to how the economy changes. The probabilities are very correlated, the expected value is rather meaningless then. It's closer to having a 50% chance to either get a full return or get zero returns, depending the macroeconomy, quite the gamble. Actually, those people were in a rather dodgy situation in the first place, or are not great at decision-making, so it might be more like 50% chance either of getting 50% return or getting 0% return.

      PS: Just elaborating on your point, not meant as a counterargument, I know you said the same thing.

    • > The problem w this model is that it only works if assets are relatively uncorrelated (it was true in the past but ignored systematic risk and adverse selection in originations). What this has to do w musk or spacex I’m still not sure

      What this has to do with with SpaceX is that there's the same blatant disregard for sound financial analysis by the very institutions that were/are supposed to know better. The NASDAQ 100 fast track decision is a similar level of financial malpractice as the ratings agencies slapping AAA on things that they knew were little better than junk. The abuses of the subprime mortgage originators were well known long before the actual meltdown. As were those systemic risks you spoke of. They were ignored by those whose entire job it was to not ignore them, and they sold out their credibility for a quick buck. If you can't see the similarities to the present situation then I can only wish you luck.

  • It is adversely selected, but it's not debt, it's equity, so price action can go real fast and nobody will be burned except folks who soberly-or-not opted into this. Everyone _knows_ Elon is the way he is, so nobody will be _surprised_ at things. No surprise, no crisis.

    • They're going to force a S&P500 index listing on IPO day so we're all going to be forced to baghold this regardless of if we want to or not unless you've got $0 in any major retirement fund.

      28 replies →

    • Friendly reminder that SpaceX is going straight to the index—Elon agitated for it. The 401k of everybody in America is serving as a bailout fund for X and now cursor, and whatever other trash he hovers up

      3 replies →

  • It’s also worth noting that Musk helped successfully lobby the NASDAQ to implement a “fast entry” rule which takes effect at the beginning of May, suspiciously convenient timing for a SpaceX IPO, so much so that I believe it has been derisively called the “SpaceX Rule”. It allows mega-cap IPOs like SpaceX to join the Nasdaq-100 index in just 15 trading days.

    Now why is this bad? Well, if you invest in a fund that is based off of the indices, you’re going to be investing in SpaceX whether you want to or not and I certainly don’t think 15 days is enough time to sus out whether this is a stable investment worthy of being in the index, but it’ll be great…until it drags a million retirement funds down with it.

  • We are better now that we learned from the first time.

    • Ug wants to borrow ten of my best sticks in exchange for future options to buy berries from his friend Og. Og has a watertight deal with Oog to invest the sticks in a five year mammoth hunting expedition but Oog first needs berries to exchange for sticks to cover his exposure on berry-puts he’s take out against Urrrg’s remortgaged stick pile.

      Well, I said no. Not getting burned that way again!

  • it's just codex and anthropic rapidly improved their AI when they opened themselves to Developer workflows.

    Google and others were sitting at the corner, laughing that they gonna burn their money for no reason! they turned out to be wrong.

    Turns out offering discounted/subsized tokens to developers massively improves your AI compared to just being a talking parrot for normal user workflow where you do not get "instant feedback" on if it worked or not.

  • Well, there are some very important differences. 1) It’s super well known what’s going on with SpaceX. Every investor should know that there’s a lot of good stuff along with some steaming hot garbage. 2) SpaceX isn’t systemic to the economy. If SpaceX and all its subsidiaries shut down and its investors got nothing back, it wouldn’t be that big of a deal.

    This type of bundling is just what conglomerates do. Is it a good thing? Not really. Many investors also hate this kind of stuff and avoid investing in these types of companies.

    • On point #2, they are trying to do that right now. If spacex is fast tracked into the indices, passive investors via index funds will be forced into buying.

Is anybody using Grok or Cursor still? I've not used Cursor since the summer of 2025 and I've never bothered with Grok for coding. Hell, I've used Windsurf briefly for a few months.

I know a ton of people that use Codex, Claude, OpenCode but can't name a single person that uses Cursor or Grok that is knee deep into agentic coding.

  • Our company (~25 engineers) uses it across the entire engineering and product orgs, and yes we are quite deep into agentic coding. We use their cloud agents for a lot of things, e.g. automated investigations of alarms, handling most customer support issues that end up hitting engineering, pre-processsing linear tickets before humans triage them, bugbot for PR reviewed with learned knowledge. Although recently they have felt like they are pulling the rug out on our legacy plan, so we may end up switching.

  • I'm on the legacy pricing annual pro plan (equates to $16/mo). I just use Opus 4.6 and have never hit more than 400 requests, so my pricing remains very cheap. That + the tab complete is what keeps me using it.

    This news is very off-putting though. I will definitely not be renewing. I don't want to be associated with Musk in any way.

  • We just started adopting Cursor over the last six months. We have at least a hundred developers using it and several more signing up every day. I'm one of them and I'm quite happy with it.

  • I have claude and cursor. I enjoy cursor. It has shortcomings but its a strong product.

  • I use grok for various subagent tasks. It's super cheap and 100tps. Never for actual thinking though.

Nobody mentioning how weird SpaceX is becoming? When it IPO it won’t be a space company anymore, but a weird whatever Elon latest ventures craziness conglomerate of some sort, plus “financial engineering” (euphemistic) shenanigans

That's a hefty payday for a model that barely functions! Every time I run out of API credits and get kicked back to Composer 2 I feel like I'm better off just packing up for the rest of the month.

I feel like we're finally at a point where you don't have to constantly argue with and constantly babysit coding models, which makes it even more frustrating when you're suddenly forced to deal with one that ignores your instructions and gets stuck in thinking loops again.

I suspect it's the vast troves of training data rather than any tech that Cursor possesses that SpaceX is after...

  • Cursor is still the best coding environment and hardness. It's actually not really close. They are so good that they actually made Gemini usable.

    The problem is they can't compete with Anthropic and OpenAI because they can't sell Opus and GPT at a discount to subscribers like OpenAI and Anthropic do with their subscriptions.

    So they either need to build a competing model or slowly die.

    • I personally disagree on the first point. Claude code in a terminal with vim is much nicer. I just don’t see the need for the bloat of an IDE when the CLI versions work so damn well now.

      2 replies →

    • > They are so good that they actually made Gemini usable

      I think Gemini is best model out there, and it's not Cursor who you should praise. I use it with jetbrains junie. Vastly cheaper than claude, faster, produces better quality code, actually listens to your instructions, more accurate. I'm sure claude code cli has some cli magic that I'm missing out on, but having everything just work in a nice IDE (and llm to actually understand your symbol table) is like magic.

      2 replies →

    • That's why I'm so puzzled to why Composer doesn't work better when they have the ability to train it from scratch for their agent harness! Yet it still fails to apply edits, gets confused why it can't call some commands in its sandbox, the list goes on...

    • They seemed to be doing fine with Kimi distillation. Not speaking from experience though, I prefer to use my editor.

  • I doubt they're buying it for Composer, I imagine they're buying it for the agent harness. It's arguably the best non-Anthropic agentic coding harness, and you get _all the models_ for one subscription price.

  • Cursor is great. I was using it up until recently. Then I switched oh my pi, and honestly I haven't looked back. I've also heard great things about open code.

  • I actually really like Composer 2. For my use case, between the planning tool, and getting it to ask a lot of clarifying questions, I regularly get very good results. I'm not doing anything complex though; mostly staying in the lane of very common web app type code.

    • It definitely feels sufficient for questions and planning, but it is surprisingly lacking in the actual coding department once you go for edits that need changes in multiple files. Which is surprising considering they should have been able to train it on their own harness!

  • Is Composer 2 a bad model because Cursor are bad at training models, or because they are compute constrained? This deal will provide the answer to that question.

Wow, we are seeing the dark underbelly of the beast here. Nobody talks about cursor anymore for a reason. Look, I'm not saying it's not useful and discounting anyone getting value out of it...

But it's clearly not worth 60B dollars in April 2026.

  • Yep. there's absolutely no way that Cursor is worth that much.

    for contrast, Elon paid $44b for twitter back in 2022. When you adjust for inflation, Twitter was acquired for $49b in 2026 money. Cursor getting bought for 1.22x more is just insanity.

    Elon seems unwilling to shake off the image that he has basically no idea what he's doing.

  • It is not cash though. SpaceX does not have $60B liquid cash instruments.

    More accurately it is 3.4% of SpaceX at the last rumored valuation of $1.75T.

    • No longer rumored as they filed for IPO!

      This is actually an amazing sweetheart deal for Cursor. Many times with these high profile acquisitions, most stock is tied to LPA's and employment at the company, and also earnout provisions. The company then finds a way to parachute them out early, which both voids the earnout and their employment, thus they never vest most of the units and the few units they do vest get bought out at 409A valuations which are typically much, much lower.

      In the case of Cursor this is an amazing boon as SpaceX listed at an almost 100x multiple which is absolutely staggering. Had SpaceX stayed private they could have 409a'd Cursor and got it for effectively ~100M$ cash.

      1 reply →

    • There's not going to be $60B of exit liquidity if/when spacex IPOs. Maybe the suckers will be banks lending against the bubble valuation.

  • The only reason I haven't switched back to VS Code is pure laziness, not using any AI features in Cursor other than resolving diffs these days.

  • Just because it's not discussed much on HN does not imply it is not relevant in the broader space. Cursor is still very much prevalent there with 1 mil DAU.

    • I’m curious if that 1 million DAU still holds as of today. I think it was reported last year some time aka before December when Claude code exploded. A quick google didn’t turn up any results that actually contained sources for the number.

Rockets, satellites, social media, AI - the only thing missing from the SpaceX hype portfolio is a certain coworking company. That would really set them up for an exciting IPO.

What are we even doing here.

I have no idea what this has to do with aerospace, but I know a bit about software and this does not look great. Cursor is obviously on a serious decline and has little to no moat in the area they are building in (IDE), which we kinda now know is maybe not even the right area (CLI). I feel like this is just a bad move?

  • Isn’t it obvious? Musk bailed out his Twitter investors with xAI. Then he bailed out xAI with SpaceX. Now he realizes that no one thinks xAI is worth the hundreds of billions he claimed it was in that potentially fraudulent transaction, and is trying to make Grok and xAI relevant by getting access to customers in the AI coding space. But in the end, it’s SpaceX share holders who are being made fools of and soon, especially with the Nasdaq fast track changes to incorporate SpaceX forcefully into everyone’s passive investments, the public will be the one who is made poorer. But Musk will become a trillionaire.

  • SpaceX is just going to be the Musk Company minus Tesla. X Corp, the X parent, is a subsidiary of xAI which is a subsidiary of SpaceX. This seems back to front, but I suppose SpaceX has the better reputation for investors whereas if X owned SpaceX the IPO would be devalued by the association with Twitter.

  • AFAIK cursor is basically the only player right now not subsidizing tokens out the ass, and has been seeing solid growth across individual and enterprise with almost every model performing best in their harness. Not sure how that’s a serious decline.

    • On the contrary, anecdotally, myself and every engineer I know have switched fully from cursor to claude code since the start of the year. I now use zed with cc. I personally could not stand the buggy mess and constant UI changes of cursor. It’s also not good value in terms of claude tokens compared to claude code.

      3 replies →

    • Wouldn’t Cursor agreeing to such a deal be almost ironclad proof they are subsidizing tokens/inference out the ass? There’s wide speculation all the large revenue growing companies right now are selling inference at break even or a loss.

    • Claude/chatGPT are not subsidizing tokens via the API and are profitable for most enterprise consumption. This meme that they lose money on every query has zero evidence and is wrong outside of the 20/200$ a month plans.

    • I would like to know where you’re seeing this, because my strong impression is exactly the opposite: a year ago, everyone was talking about Cursor, but I haven’t heard anything about it in months. It’s all Claude Code and Codex now. In terms of mindshare they seem dead already.

      2 replies →

  • Cursor seems to be pivoting to a Codex like desktop app. The real product though is probably their decently tuned harness and their composer models. I agree that their popularity has waned. I would attribute that mostly to customers chasing the most subsidized tokens and Cursor not having the pockets to keep up. Anthropic is already following suit and it seems unclear how long OpenAI is willing to continue. I think in a case of a market correction that forces model makers to adopt more reasonable growth targets, that Cursor is decently positioned.

  • Anthropic, OpenAI, and Google are all investing heavily in their desktop type apps, I think the TUI phase is coming to an end.

  • ai trends seem to mirror general coding/software trends but compressed. People used to edit programs with sed, but the ide proved to be more powerful from every perspective. cli tools always have their place for "power-users" and other specialized intermediate usecases like tui's, but in general the ide has overtaken every aspect of cli use and many devs hardly ever use the terminal. I suspect a similar thing will happen with ai.

    • While surely someone has done human-driven editing with sed, that's not what it's for. Remember that ed is the standard editor.

  • > no idea what this has to do with aerospace

    SpaceX is no longer SpaceX per se, but SpaceX-xAI.

    My TL; DR (and this is mine, personally) is its mission has pivoted from colonising Mars to building a Dyson sphere. Space-based datacentres are a demand excuse for putting lots of solar panels in space. Going one level down, more Cursor use is a demand excuse for putting lots of datacentres anywhere.

Every time Musk does anything these days, it further reveals the shell game he's playing with his companies. This is going to be an Enron type of story eventually. I truly wish I had a choice to pull my tax money out of this particular subsidy.

  • Enron was absolute peanuts compared to the financial fraud Musk has been executing (with the apparent blessing of the SEC). At its peak Enron had a roughly $70B market cap, TSLA is currently sitting at $1.74T. We can expect similar numbers from the SpaceX IPO.

    • It's hard to compare these numbers directly since valuations have increased quite a bit since a quarter century ago. As a proportion of the S&P 500, Tesla (2.3%) is about 4x of Enron at that $70b (0.6%).

  • Tesla is profitable, as a matter of public record. And SpaceX is, by all accounts, extremely profitable.

    • SpaceX is _not_ profitable by most reasonable measurements of accounting. If you discount rocket depreciation costs and R&D, then yeah its profitable from starlink revenue.

      38 replies →

    • It is less about profitability and more about dilution of ownership. He seems to have a pattern of diluting the ownership of his profitable companies by folding in his less profitable/failed companies. You still own a share of a profitable company, but a smaller share, to his benefit.

    • How much of that profit was due to public subsidies of the sort that he killed for other companies but not for himself during his tenure as a special government employee?

    • SpaceX was surely more profitable before it was used to bail out Elon's xAI which was used to bailout his purchase of Twitter.

    • just because a bunch of rockets went up without blowing up, does not mean they are profitable. it cost money to shot rocket, and it is very expensive, reusable or not. most launches are internal launch without external paying customers.

    • Have you looked at their latest report?

      They are only profitable because of subsidies. Pretty much 1:1.

    • Genuine question, how do you know that without a 10K? Have the filed any document that shows their finances?

  • Matt Levine writes a bit about this - the Elon Musk Mars Conglomerate. And really if you're investing into e.g. SpaceX you're not investing into SpaceX you're investing into the Elon Musk Mars Conglomerate. And most people seem to want that.

    Tesla's the odd one out: it's public but it's still in there, although Musk would probably prefer it to be private too.

    • Tesla is the free cashflow play that is probably the most important for mars as there is no distilled fermented dinosaur juice on mars, but considerably more by ratio of lithium / oil than the Earth. Our flintstone fire mobiles won’t work so well there, and battery / solar will be important there for everything, including mobility and armies of slave robots.

      4 replies →

    • > Elon Musk Mars Conglomerate

      That’s SpaceX’s version of Tesla’s self driving car pipe dream

      Edit - I use self-driving car and Autopilot interchangeably

      11 replies →

  • [flagged]

    • Words do mean something, and you could have taken 5 minutes of research to make a reasonable counterclaim

      Tesla has an insane PE ratio because it’s a casino stock (~350x). As a comparison, NVIDIA IS 40x. SpaceX Is projected to be 300-500x. These are fantasy, completely unrealizable valuations. Similar to Enron, and Enron was over 70x. Enron wasn’t some surprise either.

      Typically when PE gets out of whack, market analyzers dig into what is happening because it’s usually chicanery. No longer. Everyone is along for the ride.

      2 replies →

    • > How are any of these companies at all related to Enron?

      There's a lot of parallels:

      * Circular transactions between companies under the same control

      * Using SPVs to keep debt off the books

      * The supplier funding its own customer through investment to inflate revenue on both ends

      * Valuations driven by a hyped up narrative and decoupled from actual fundamentals

    • Tesla isn't that profitable, but SpaceX is likely generating boatloads of cash. From what I can tell Starlink alone has a free positive cash flow of about $2 billion. I'm not sure what the launch business is worth, but it's likely a lot given the absence of domestic competition.

      I have a suspicion the reason Musk wanted to combine SpaceX and X.ai is the latter gives him losses to write off against all that cash from the former plus a chance for a big AI payoff.

      7 replies →

You can hate Elon or just be misguided about deals in general. This is brilliant. He’s buying revenue and, on the thesis of scaling agentic knowledge work replacement, a user of his GPU clusters and ultimately GPUs in space. A $60B option is a premium on their revenue - but it may look cheap if he accelerates their coding models. For Cursor, they get what’s nearly impossible to come by - real capacity guarantees and de-risking their reliance on Anthropic or OpenAI.

Laugh all you want. He may have the last laugh on this one.

Elon doesn't know what to do. Ani failed, no one apart from his alt accounts is interested in Grok pictures.

Since the firing of several Grok founders, Grok has decreased in capabilities. It is illogical and insults users when called out.

So he does what everyone does. Write more dev tools, slap a price on it and hope retail investors will be impressed in the IPO. The $60 billion is of course optional and will just be used in the IPO to inflate the valuation.

Wow. Tech CEOs and investors have completely lost touch with what money really is worth.

How is a VSCode fork and a open weight LLM fine-tune worth $60B?

One would think Elon would learn his lesson after overpaying for Twitter and then having to merge his failures together to stay afloat. But no, more cash into the burning pile.

  • > How is a VSCode fork and a open weight LLM fine-tune worth $60B?

    Ignoring future business ideas, Cursor reported reached $2 billion+ annualized revenue run rate in 2026, doubling from 2025. Recent financing rounds reached high-end valuation between $30 billion and $50 billion.

    • Revenue without expenses is meaningless. Annualized revenue is even worse. It's like a gambler bragging that they spin through $20,000 a month. Yeah, but for how long?

      If you give me a billion, I can do an annualized revenue run rate of ~$12 billion just by selling a dollar for 99 cents.

      1 reply →

    • “annualized revenue run rate” is a bogus accounting term. It’s like taking a paycheck and multiplying it by 365. Notice the complete lack of any mention of profits.

      5 replies →

    • To add to this, Cursor provides a high value go to market strategy for X.AI's modeling efforts. Cursor's own modeling efforts would require an extreme investment of capital to compete. Capital which X.AI has already spent or is planning to spend.

    • I don't think we can use normal valuation methods for these AI companies.

      Things are moving so fast, and these companies have no moat whatsoever. Purchasing a company for 30x annual revenue (and as others have pointed out, how much of this revenue goes straight to companies like Anthropic?), without knowing if it's even going to exist in 3-5 years, seems bonkers.

      I mean, congratulations to the founders on becoming billionaires in record time, but this is uncharted territory.

  • I don't think it's about worth any more at this point; it seems more about money laundry and manipulating the market. They are shifting power between each other and create an illusion of a healthy economy, not carrying about the damage they create for everyone else.

  • Yes, and Whatsapp was just a messaging app with a stupid Erlang backend. These deals are not about the tech, they buy the business, that includes the brand and the user base. Whether we think it's worth that amount is indeed up for discussion.

    • Whatsapp was much much more at that point. It also had a huge userbase at a time when getting such a number of people was incredibly difficult. Many were also paying the $1 per year fee. Switching from Cursor to Kilo etc. takes nothing. There are no "friends" you need to convince to switch.

    • Yes. But unlike cursor, Whatsapp had the following advantages:

      1. It's cheap to run.

      2. It has clear advantages over existing technology (SMS).

      3. My mom uses it. She's never going to use cursor. Whatsapp had a huge userbase in Europe. Basically everyone I know uses it.

      And it was "only" ~$20 billion. Inflation can't be this high.

      8 replies →

    • WhatsApp had real network effects built in, and network was the moat. Don’t think Cursor has any real moat.

    • The cost of switching from WhatsApp to an alternative is huge. You will lose access to your family, friends, whomever you chat with. They need to switch with you and in turn their friends need to switch with them, and so on.

      The cost of switching from cursor to codex or Claude code is minimal.

      So what does Claude code actually have that spaceX can't imitate? Well, not much, but acquiring hot companies before your IPO is a good strategy to drive up your own valuation.

  • Let's face it - Grok is not nearly as popular among programmers as Claude or Codex, and that means that xAI is not able to vacuum all the data that his competitors have access to.

    Cursor is installed on a LOT of computers.

    Once Grok becomes the default engine, it will raise adoption.

    More importantly, if you have Cursor installed all your data may be sent to their labs whether you use it or not (unfortunately - this is par for the course for all the LLMs, a la Microsoft).

    That's worth a lot - especially considering that Cursor might also grow with the shift to more powerful local models and the fact that it has a respectable income stream.

  • > How is a VSCode fork and an open weight LLM fine-tune worth $60B?

    Corporate contracts. A lot of companies have signed onto Cursor. xAI has a pretty toxic brand with Elon and the nonconsensual sexual images scandal. xAI has a ton of compute and few corporate customers. Now they have a ton.

    > One would think Elon would learn his lesson after overpaying for Twitter

    I think he took over Twitter to control what people using it see and promote right wing viewpoints. To that end it’s been a wild success.

    • > xAI has a ton of compute and few corporate customers. Now they have a ton.

      For now, perhaps. I work with numerous companies who refuse to do business with his brands.

      1 reply →

  • > Wow. Tech CEOs and investors have completely lost touch with what money really is worth.

    Remember those stories of lottery winners who win millions then wind up broke AF, even homeless, in only a year or two because they have no impulse control and blew through it? Same people.

  • It's purely about control, not profit.

    He bought Twitter to grab a megaphone to propagate his ego/agenda.

    Grok was built for similar reasons.

    He's buying Cursor to have a tool to push Grok on the world, and to have something of his own under his own ideological control to compete with SamA and Dario.

  • It's not the codebase they're purchasing, it's the customer list, R&D talent, and most importantly, data corpus.

    X AI is weak on coding and most of their founding researchers have departed for greener pastures. Musk has a great datacenter, but no researchers or data to use it with. This acquisition makes sense in that context.

    See also ($, Stratechery) https://stratechery.com/2026/john-ternus-and-apples-hardware...

  • You have to look at it now like cryptocurrency “market cap” numbers. It’s more of a marketing tool than anything. This is just another honking whirlygig to bolt onto the SpaceX IPO to try and generate exit liquidity

  • he doesn't act like he regrets buying Twitter/X/Xitter

    maybe he's getting value from this? (also the deal was essentially secured with Tesla stock, so who knows what did he actually pay)

  • > How is a VSCode fork and a open weight LLM fine-tune worth $60B?

    The same way a rocket company that counts short-lived as satellites as an asset is worth 1.7 trillion. Congratulations to the Cursor folks, they are the only winners in this.

    This is also part of the AI bubble delusion: agent assisted coding works, at least for some people, for some purposes. This deal perpetuates the illusion that AI will find high value use cases. The reality may be that software development has unique characteristics you won't find in law or medicine or other domains.

  • Buying Twitter played a key part in getting Trump re-elected, so I think Musk figures he got what he wanted in terms of deregulation, dropped prosecution, and damage to his political opponents.

    This deal is different: SpaceX is heading for an IPO which is now complicated by xAI becoming a subsidiary. Cursor is actually popular and I’m sure this is all stock-based so as long as investors believe that those users stick to xAI it’ll juice the entire SpaceX IPO. I am skeptical but these days the market seems to be driven by a country-club full of guys in Connecticut who are constantly hyperventilating on X so maybe from that angle it’s just another way he’s getting what he wanted from Twitter.

    • Quoting a viral tweet: “Elon is such a dumbass, he spent $44 billion on Twitter and all he got was control of all 3 branches of the federal government.”

      2 replies →

    • Buying Twitter played a key part in getting Trump re-elected Some would say it was simply caused by the mediocrity of Bidden and more globally of the democrats. You don't need to brainwash people for this.

      2 replies →

$50bn for a harness makes no sense, what am I missing?

This seems like diworsification to me. Cursor has nothing to do with rocket launching, so it will be a distraction for management.

Dammit, I liked cursor

  • same. i finally tried Claude Code and i just shrugged. Cursor definitely has a clunky UI with an identity crisis, but it pioneered plan mode, and auto / composer chugs along without rate limits for the most part.

    Opus, I watch my allotment creep up every turn…

It never occurred to me that LLMs would be used in the development of something like rockets and space-going vehicles..

A sobering thought.

  • Yeah, maybe it'll sober some people up to stop pretending they can't see how useful LLMs are in pretty much everything. A sharp tool to wield, easy to cut yourself with, but also extremely useful.

    Honestly, it felt much stranger to me to learn, a few years ago, that they're 3D-printing rocket engines. With my experience limited to building my own PLA/ABS 3D printer out of salvaged motors and parts printed on another printer, it was hard to imagine how this is anywhere near safe and precise enough. But turns out, FDM-ing some plastic blobs is not the same as fusing Inconel powder with lasers. Same with using LLMs for software engineering (whether in aerospace culture or otherwise), it's just not the same as asking ChatGPT "please make me an app to do something idk how i cannot code send halp".

  • xAI also owns Twitter, which is even less useful for SpaceX.

    They need xAI as a reason for the narrative that data centers will be in space, so SpaceX can project far more growth before the IPO. After the IPO they'll find out that data centers in space are too expensive and overheat.

This is a classic Elon move. He bundled up his company that is, shall we say, crap, into his most valuable company, then tried to hype it up as much as he could. Like when he promised Tesla cars would self drive in X years but it never happened, then pivoted to AI/robots, then re-routed Tesla’s GPUs to xAI, etc.

Cursor might not be the new hotness, but if we believe that agentic coding is the next wave and we’ve gone from asking chatbots to actually using agents for coding, then yes, this move makes sense for Elon to hype up a SpaceX IPO.

I feel Cursor isnt’t even worth $6B. What is the moat, the value, the sauce here?

The “apply” model to turn LLM output into code changes?

I like SpaceX a lot but this really doesn’t make sense at $60B

I really don't know what Elon is thinking here because SpaceX's IPO is already precarious, for several reasons:

1. It was used to rescue himself and key investors from overpaying for Twitter, which was first rescued through xAI (and I don't know why anyone thought investing in xAI was a good idea but here we are). If our regulators weren't defanged, this deal would've gotten alot more scrutiny (IMHO). Whatever the case, this is all diluting the SpaceX business for overpriced AI vaporware;

2. From what I can find, SpaceX's revenue in 2025 was ~$18B with a $5B loss. That doesn't sound like a $1.5T+ company to me;

3. The markets are being rigged to make the IPO a success by changing the rules to force passive funds to buy into it with a tiny float (5% instead of the normal 25%); and

4. Here's the big one. I think Starship is a badly designed program that's going to take many billions more to complete and commercialize. There's not really a market for bigger payloads (evidence: ~1 Falcony Heavy launch per year) and STarship will effectively have to compete with Falcon 9 at a time when reusable alternatives (eg from the Chinese as well as Blue Origin) are coming to market.

  • Starship won’t compete with F9, or BO because it is fully reusable and cost less than either. The Chinese are not a player in the global launch services market at all so don’t count.

    Starship isn’t comparable to Falcon Heavy because it has vastly more volume, which will make it the cheapest way to launch Starlinks, which will be a lot of launches to begin with.

    • Starship is absolutely competing with Falcon 9 in two ways:

      1. If F9 is cheaper, why would you use Starship?

      2. If SpaceX decides to force Starship adoption by simply halting F9 launches or making them prohibitively expensive, well the market is still open for an F9 clone from someone else.

      My point with FH is that there isn't a demand for much bigger payloads. Now, SpaceX wants to induce demand with STarlink launches. OK, so the viability of Starship is tied to the viability of Starlink. I thought Starlink was a clever way to prove F9 reusability but the first-mover advantage won't last forever.

      My main point here was that F9 was developed, SpaceX's competition was the likes of ULA with their insanely expensive rockets but whether SpaceX uses F9 or not, it has become the new baseline.

      SLS is insanely expensive to build and launch but it still has a much greater launch capacity. Starship's solution to this is essentially in-orbit refuelling where, I believe, it will take ~10 Starship launches to refuel a Starship in orbit. This too is an insanely complicated capability that they haven't even begun to develop yet.

      And whatever the ultimate per-launch base cost works out to be the program R&D cost has to be amortized cross those launches so even if the base cost is $10M, if you spent $10B developing the program, it still matters if you launch 10 a year vs 100.

      When SpaceX goes public, they're going to be forced to disclose a whole lot more information about the program cost and I suspect it's going to be a lot higher than the rosy projectsions you get on Twitter.

      Another way to put it is I think Starship could be SpaceX's Cybertruck.

    • > Starship won’t compete with F9, or BO because it is fully reusable and cost less than either.

      Currently, this is not the case. Not fully reusable, and not costing less than F9 or BO.

    • There is a lot resting on Starlink, 11 gigadollars in direct revenue that accounts for fully 60% of SpaceX's total revenue of 18 gigadollars. It's hard to see how that level of revenue can sustain a 1 terradollar valuation.

      Like, TSLA had 94 gigadollars in revenue last year, and it's a 1.2 terradollar company, and most outside analysts are frankly skeptical of that multiple. SpaceX is trying to get a similar valuation on a fifth of that revenue.

This is the right partnership to happen. SpaceX has all the compute but is missing the talent for training LLMs, especially on the RL side. Cursor has the talent and RL stack, but doesn't have their own pretrained base model or own their compute. Both will be on a bad trajectory without cooperating because Claude Code and Codex have gained so much momentum already.

“ Cursor has also given SpaceX the right to acquire Cursor later this year for $60 billion”

That isn’t an agreement to buy

I am part of a discord group with about 1000+ devs. I polled them in Jan to see if they had dropped cursor for claude code.

80% of those responded (250ish in the group had). Bit of selection bias there from the question - but my impression was Cursor is very much dying to competition from the labs.

If SpaceX grabs cursor and then gets its models into use for a huge amount of companies and developers, this might actually be a very wise move. So I don't think SpaceX would only pay for the harness, but to access to a mass of potential users.

I am so actually beyond sad that I ever trusted Musk, all the signs were there, from the lies with Tesla to the nonesensical point to point "tourist" lies, to the Mars lies, to the fact that the spaceship they are developing right now requires an actual elevator to get astronauts down, it was never meant for humans, it was meant to deploy sats in space even cheaper, outcompete the competition and basically kill human spaceflight as a result... because less profitable human rated spacecraft won't be viable.

Oh yeah, did I mention how Starlink is literally already in the close to Kessler Syndrome territory? all it would need is for a strong enough solar storm to hit their sats.

  • The elevator was there when it was originally announced.

    There's no Kessler Syndrome where Starlink is.

    You'd know this if you read Kessler's first paper. It's online.

    i.e if every single Starlink satellite crashes into another you won't get Kessler Syndrome.

    And the same it true for the planned Kuiper.

    • First of all, yes I know about the elevator hence why I mentioned it, you know, first of all it's not that safe to be going down an elevator from what is basically a multiple stories high building while in space (#1) and (#2) why would you add complexity/failure points on purpose if your mission was being multiplanetary?

      The spacecraft wasn't designed with humans in mind first.

      And second:

      This is a paper by Kessler himself:

      https://conference.sdo.esoc.esa.int/proceedings/sdc9/paper/3...

      1 reply →

Could a mod please consider changing the story link to one of the actual articles shared in the post? I don’t think we should be giving the white nationalist, non-consensual-porn-generating social media site any free traffic.

ITT: The same geniuses that predicted with certainty X will fail are also predicting, with much less certainty, that "Oh God, let this be the end of Musk"

Wish I played that interview game better. I saw the success coming from a mile away (2022) but I can't vibe with people in the hire game right. It's like eye contact, smiling, facial expressions, stuff like that.

I guess there's a bunch of tools to not suck at this. Anyone had success here? The AI tools say I'm great because they can't pick up the kind of problems I'm talking about.

  • Pretend to and/or be motivated by things other than money, that’s the strongest thing interviewers drop people from, even though they’re motivated by money to be there.

    • Interesting. I genuinely do not care about money.

      The motivation of money is literally zero to me. Maybe that's a problem as well: they want people who Are motivated by money acting like they aren't?

      I wanted in because I saw them doing exciting impactful things That's literally it.

      I dunno. I've been struggling with this for decades

      2 replies →

>acquire Cursor later this year for $60 billion or pay $10 billion for our work together.

This seems like an elaborate Elon rug pull. A Windsurf situation 2.0

$60B for a VSCode fork with AI integration... It may show the value of the gap between vanilla LLM output and production-ready applications.

Misleading title on the post - SpaceX has the OPTION to buy them for $60B later this year, or pay $10B for their work together.

I’m no fan of Elon Musk, but even from a neutral perspective I’m bewildered by the merger between X, X.ai, and SpaceX and now this acquisition of Cursor. What’s the endgame? How does this help with the whole vision of “we all live in space and mine resources from the Moon and have data centers on Mars”?

  • The endgame is to game the index funds by bribing or otherwise.. convincing the big stock exchanges to forgo their index inclusion rules so SpaceX will get included in Nasdaq 100 within 10 days or something stupid like that. SpaceX will initially float a tiny fraction of shares at a wildly inflated value and use a combination of artificial scarcity and Elon Stans (retail) to keep the stock from crashing until it gets included in the indexes. Then, your 401k will auto-buy SpaceX, letting insiders exit at their ludicrous valuation.

    Eventually, stock prices will correct hard, and retail/passive investors will be left with the bag.

    The idea that merging these companies has some business purpose is hilarious. It’s purely financial engineering. Unfortunately, our existing system has little consumer protection against this kind of fraud, so Elon will probably get away with it, at least in the short term

  • idk but feels like this might be a new literal kind of acquihire, to bulk purchase workers in cash

  • The Elon Musk Company does what Elon Musk wants. Tesla is dying, X is a disaster, so he bundled everything into the one company that had a bright future, SpaceX. There is no grand vision or endgame beyond do as Elon Musk wants. Going to Mars or the Moon or whatever was never a vision or mission, just a story to tell.

    • Model Y is still the best selling car in the world (and still the best-selling car in China), but yeah Tesla is *dying.

  • > What’s the endgame? How does this help with the whole vision of “we all live in space and mine resources from the Moon and have data centers on Mars”?

    I put this in the other thread, but my personal working hypothesis is the SpaceX/Musk mission has pivoted from colonising Mars to building a Dyson sphere.

    Space-based datacenters are a demand excuse for putting solar panels (and eventually, solar-panel fabrication) in space. Cursor is a demand excuse for building more datacenters (and eventaully, learning to fabricate chips). If I'm correct, the next acquisition will be in some chip or solar-panel fabrication bottleneck.

    To be clear, I'm not advocating for this mission. (Though I do think the space-based datacenter pitch has gotten more scepticism than it deserves. For realistic interest-rate and terrestrial-delay values, assuming ongoing AI demand, it breaks even for surprisingly-proximate radiator-mass values. Obviously more problems beyond my toy model to solve. But I expected the math to say fuck you out the gate.)

    I'm saying this is a good working theory for explaining–and predicting–Musk and SpaceX's actions. Mars explains why SpaceX's engines burn methane. Dyson sphere explains why xAI is building massive datacenters and now finding acquisition targets to fill them with.

    • > building a Dyson sphere

      So they are trying to take everyone’s money in bigger and bigger chunks until there is no economy left but hype.

      Folks, if we spent 1/10th of the time and money we spend on this bullshit on taking care of people’s basic needs and education we would be far closer to the sci-fi future everyone seems so motivated to get to. Covid and the Trump cult seem to have broken almost everyone’s brains and we’re all gonna pay for the hubris.

Random data point: as a long time VSCode user when I first heard the hoopla about Cursor I rushed to try it. Didn't work (at all). So I added my name to the open bug report, waited a few months. Tried again. Still didn't work. Became a Claude Code user and never looked back.

It looks like this is just an "option" to acquire Cursor at that price? Implying they only plan to exercise the option under certain conditions (such as, one might presume, Cursor actually being worth that much. As right now it definitely isn't.)

I know Cursor is getting economically not so viable compared to OpenAI and Anthropic offerings but with a deal like this they could also offer $200/mo plans that are attractive. Obviously _if_ their models are good. We have to see!

I remember using Cursor earlier this year and it was so unoptimised for average machines. An LLM provider extension on VS Code is so much better. I don't get why is Cursor still around. Is it because it was first to market?

I guess it makes more sense than shoe brands pivoting to GPU provider.

Spacex already owns Twitter and xai, trying to post-rationalize with justification like they have servers doesn't make a whole lot of sense. It's all accounting at this point.

Welp this just removes them if they get bought (and likely also if they coordinate even more with an Elon company) from my list of tools I’ll use.

Crazy a fork of vscode is worth 60B. What’s vscode worth to Microsoft? 200B?

xAI is working on virtualizing white collar workers. I'm guessing this is part of that.

See also: companies buying up the Slack and email archives of defunct startups, for training data.

Hard to know whether development will remain an activity that lives on a local machine for much longer.

This could be a lot of money to spend to acquire users that may not be sticky.

Surely you mean xAI right, surely it's a typo? Right...?

The same "mistake" that SpaceX bought 10% of Tesla CyberTruck?

Wait are they all Musk's companies? Is it a pattern?

/s obviously

I wonder if they are actually 'acquiring' some of the existing contracts between Cursor and X/Y/Z rather than the product itself.

I thought the world was agentic vibe coding. No idea needed. If this world materialized, then there would be no need for an IDE.

I'd be interested in this breakdown - what % of that is cursor's product(tech x customer) vs future tokens

Last day for me using Cursor at work, I prefer to move to Codex and Claude Code that touch anything related to Elon.

Gross. We need more anti trust enforcement. Large incumbents killing all competition will make us weaker over time.

SpaceX, xAI, Collosus data centers, next space compute, X, Starlink and soon Cursor to join 2, 3 & 4 together?

Complete waste of $60B. It's just a prompt+tools. This is how you destroy SpaceX from the inside.

This looks like SpaceX playing catchup to Claude and OpenAI that already provide coding solutions.

60B for Composer 2…that is built from Kimi K2… what ever happened to “Grok being the best”?

  • Am I the only one that thinks Composer is really good, when you factor in the speed and the cost?

    • I don’t doubt it is. End of the day, it’s a fine tuned Kimi. They tried to hide it and making their work sound more impressive than it is. It’s easy to have stuff be cheap when you don’t have to train your own model from scratch.

    • Composer is clearly dumber than the rest but then I only ask it dumb questions and it answers them really quickly.

You sour pusses are wrong. This is a smart move that amplifies a brilliant team from cursor with serious compute, raising the odds Elon can get to the frontier, which is worth so much these numbers will all look like a drop in the bucket.

I imagine none of us had this on our bingo cards.

If this is an acquihire, it doesn't compute for me (though I can't say I understand how things work in the world of the 60B level). LLMs are new enough that nobody has a big enough headstart to warrant a 60B personnel change.

The IPO angle also doesn't make sense. Musk cultists were gonna buy anyway; this doesn't change that. And for everyone else, who wants to pay down debt on an acquisition whose effect will almost certainly not be palpable in mainstream circles, if at all?

I don't fully understand the influence that comes with SpaceX subsidies and government contracts, but I gotta believe that rounding up non-lab AI chops are on that agenda?

The exact options - 60B for acquisition (obviously not a cash deal, right?) or 10B for unspecified services rendered... also don't make sense for either of the first two.

Is this just a way of the government securing contractors by proxy that wouldn't pass muster if done through the normal channels?

You sour pusses are wrong. This is a smart move. Cursor has a brilliant, capable team with serious model chops who will be able to boost the odds of AGI success. They also come with a revenue generating machine.

I really don't think Cursor is going to be acquired for $60 billion. That price is absolutely absurd. I agree their harness is excellent, but it's hard to argue they have an overwhelming competitive advantage over rivals like Claude Code and Codex, or open-source alternatives like OpenCode. What's left then is Cursor's data, talent, and user base — but even accounting for all of that, the price is still ridiculous.

I've personally watched a lot of developers around me (myself included) who were enthusiastic Cursor users when it first launched gradually migrate over to Claude Code and Codex. And I don't think this is just happening in my bubble.

My guess is this is some kind of strategic play ahead of SpaceX's upcoming IPO — an attempt to get a higher valuation stamped on the company. But I'll say it again: $60 billion is absolutely absurd.

Well I am glad I built my own IDE now so I can switch off of cursor and don’t have to participate in training the model of an aspiring monopoly.

DM me if you want an invite. I am keeping it to a small on purpose.

I think this is great and helps x.ai building Grok Code and Grok Computer.

It is good to have more competition in this area.

So there aren’t just 2 big players which also have their ideological flaws.

SpaceX’s announcement (non paywalled):

https://x.com/spacex/status/2046713419978453374

  • "SpaceXAI and @cursor_ai are now working closely together to create the world’s best coding and knowledge work AI.

    The combination of Cursor’s leading product and distribution to expert software engineers with SpaceX’s million H100 equivalent Colossus training supercomputer will allow us to build the world’s most useful models.

    Cursor has also given SpaceX the right to acquire Cursor later this year for $60 billion or pay $10 billion for our work together."

is this just to drive up the buy price for others while having no intention of buying it themselves?

reading this thread, I seem to be the only cursor user on earth on the free tier using tab-completes.

Good, I needed a reason to cancel my Cursor subscription.

I associate Musk with being user hostile, unreliable, meme oriented and disruptive in the worst sense; I’d like my work tools without that please.

I think cursor is worth literally $0.

It is a complete rip off of someone else’s work vs code and is literally no better than copilot.

Makes me mad it’s even allowed to be a commercial company ripping off vs code.

It is a completely idiotic purchase. Literally one person could recreate the entire IDE themselves… like it’s so easy to make.

And there’s no brand loyalty to it because people want the best llm they don’t care about the little plugin so no moat

but What exactly is SpaceX doing in the AI Space (Pun Intended) and Why?

these are weird times...

Can't wait for this idiotic bullshit bubble to burst.

A rocket company buying a so so overvalued coding AI company is a joke even worse than the 2000s internet pet food companies were

Looking forward to seeing where this goes, both companies have a reputation for engineering excellence.

shoe company goes AI, rocket company goes AI, it is market signaling.

  • That would make more sense if SpaceX hadn't absorbed xAi just 2 months ago. The rocket company already went AI. The signal was sent already. This is just a bad business decision.

Absolutely retarded and absurdly overpriced for a tool that's basically fallen out of use. They're just trying to do anything they can to justify a crazy IPO valuation so they can keep pumping money into xAI.

Of all Musk’s “companies” this makes least sense to me. xAI is at least in the AI space, X.com is a true tech company and Tesla is also closer to AI than SpaceX.

Guess SpaceX is the only one with the money “available”.

  • You're forgetting that xAI and X.com have both already been folded into SpaceX (First xAI acquired X.com, then xAI got acquired by SpaceX, both mergers were all-stock acquisitions so they were done with funny money). So when people say "SpaceX" now that does encompass both xAI and X.com as well. The reason Tesla wouldn't do this is because Tesla is a public company so it's more difficult for them to do insane shit without being sued.

Every now and then I feel like I got a handle on AI and it makes sense. Then something like this happens and either my whole concept of it is wrong, or this makes no sense and is insane.

Is is me or the world of finance is going crazy. Or maybe it has always been.

SpaceX, a rocket company owned by Elon Musk bought xAI, an AI company also owned by Elon Musk for... reasons. Don't give me the datacenters in space narrative, we all know it is bullshit.

It is then buying the option to buy a company for which the only contribution is a glorified VSCode plugin and the reselling of other companies LLM services at an absurd price. I understand that it is more complicated than that but 60 fucking billions, that's the GDP of a small country!

And now, Elon Musk intends to IPO SpaceX, which means he expect people to buy into all this bullshit. And considering that unlike me and judging by his wealth, he seems to be really good at understanding the market, so he is likely to be right.

Does that mean code that astronauts' lives depend on will be vibe-coded slop? Nothing is too insane any more these days.

Ohh it’s not an acquisition, it’s right to buy later for $60B or we a work together for $10B. Huh?

really happy for the Cursor team but at the same time disappointed that the biggest non-lab AI company couldn't exist on their own.

shows how intense the power laws are around ai and how much of a capital game it is.

Okay, now how do I cancel/refund the remaining portion of my pre-paid year subscription? No way in hell I will support a company owned by Elon Musk.

Question for Musk hating people - I understand why you hate Musk, but why is doing business with Altman or Microslop any better?

  • Not much better but at least they're not speaking at far right rallys and lending support to fascist parties across Europe.

    • What really puzzles me is how years of woke insanity are forgotten / forgiven, but a nazi salute is not. Remember how Microslop employees used to start their presentations with a list of Native American tribes who owned the land their office was at? Maybe people don't read Orwell anymore... that stuff was straight out of 1984.

      I see being downvoted on my question already - can people who hate Musk not see the difference between asking and supporting?

      2 replies →

Time to delete Cursor then. I refuse to support someone that is doing so much active damage to democracy and cut funding for some of the poorest people on the planet.

Government subsidized purchase of a private company. Fantastic. All funded by the taxpayer to send rockets to a dead planet and to burn up all the energy on our alive but suffering planet.

  • Point to any government subsidies for SpaceX - or do you think your salary is a subsidy and everything you do at work is worthless to your employer?

    SpaceX has paid for contracts to deliver services to the government and those services have saved the government billions of dollars compared to the alternative.

    • You're claiming that SpaceX has not received governmnent subsidies, grants, and contracts?

      > those services have saved the government billions of dollars compared to the alternative

      Source? All I can find is some guy saying it.

      And it doesn't really matter what they've saved. It doesn't excuse conflicts of interest.

      3 replies →

The other day my colleague asked Grok:

"Please estimate Elon's IQ based on his timeline"

It estimated 115-130. A decision like this points to the lower end.