[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vm / vmg / vr / vrpg / vst / w / wg] [i / ic] [r9k / s4s / vip] [cm / hm / lgbt / y] [3 / aco / adv / an / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / pw / qst / sci / soc / sp / tg / toy / trv / tv / vp / vt / wsg / wsr / x / xs] [Settings] [Search] [Mobile] [Home]
Board
Settings Mobile Home
/g/ - Technology

Name
Options
Comment
Verification
4chan Pass users can bypass this verification. [Learn More] [Login]
File
  • Please read the Rules and FAQ before posting.
  • You may highlight syntax and preserve whitespace by using [code] tags.

08/21/20New boards added: /vrpg/, /vmg/, /vst/ and /vm/
05/04/17New trial board added: /bant/ - International/Random
10/04/16New board for 4chan Pass users: /vip/ - Very Important Posts
[Hide] [Show All]


[Advertise on 4chan]


SpaceX buys Cursor for $60 billion

https://x.com/SpaceX/status/2046713419978453374
>>
i paid $0 for vscode fork lmao, who uses text editors anymore
>>
>>108656290
thanks for the urgent update and twitter repost
retard
>>
>>108656385
Retard.

>>108656290
The world doesn’t make much sense anymore.
>>
>>108656290
Yuck. I am now uninstalling before the jeetification
>>
>>108656475
>before the jeetification
What do you mean "before"?
>>
So are they going to kick out OpenAI, Anthropic and Google and just use Grok?
>>
>>108656504
Jeetier then
>>
After all is said and done I think the AI-skeptical programmers will feel a sense of relief for their caution.
>>
Did I miss a major development or isn't Grok just another probabilistic, transformer-based LLM?
>>108656537
They could ride this bubble into the next generation, until nobody alive even knows how to write code by hand and everyone relies on "AI" for everything.
>>
>>108656672
It's the same as the others, just worse.
>>
>>108656290
>$60 billion
this shit isn't sustainable.
>>
>>108656716
Money isn't real. q
>>
>>108656512
More likely the other providers cut them off.
Cursor did post train some chink model.
Elons mostly buying them because they have a mountain of training data and some of their staff.
Xai is basically dead rn (I think they're actually renting out compute).
I don't think this saves them in the long term though because no one wants to work at the child porn factory.
>>
>>108656290
Never used Cursor, but I guess it's going to be completely unusable now.

>>108656716
Google paid $2.4B to acqui-hire Windsurf's founders. AWS as able to create Kiro in-house. Cursor's own coding LLM is largely for doing simpler tasks where paying Claude Opus token prices don't make sense. SpaceX is massively overpaying because they need a big A.I. story before it goes public in June.
>>
>>108656732
until everyone asks for it back
>every market crash in history.
>>
>>108656290
ChatGPT all the way.
>>
Making an advanced inhouse AI is actually the next step towards saving me production time in game development.
>>
>>108657119
It sounds like alot of these companies are going bankrupt and other ai corps are buying them before that happens to keep the moral for the pyramid scheme high.

>>108657131
Shouldn't your game already have ai?
Don't think about just putting it into development, go the extra step and actually put the ai into the game.
>>
why does elon musk pay 20 times the value for everything?
>>
>>108657422
I imagine he gets some of the money back into his personal account for spending that money the investors gave him.
>>
>Give us $10 Billion dollars or give us $60 Billion dollars

I just don't understand how business works any more
>>
>>108657479
just jewish things
>>
>>108656290
>SpaceXAI
wait when did that happen, how the fuck did I miss that?
>>
>>108657512
a month ago
>>
>>108657516
grim
>>
>>108657512
X was not profitable so it was rolled into xAI, which had a better financing story.
Then xAI did not become profitable fast enough, so it was rolled into SpaceX, which has a better financing story (the upcoming IPO). Maybe xAI would have been profitable eventually, or maybe not, but doesn't matter now.
IPO nonwithstanding, the US government is reliant on SpaceX, so it will not be allowed to fail. Thus anything that is not making money (or not enough money to keep investors happy), Musk can roll into SpaceX and keep it going that way.
I said "IPO nonwithstanding" but there IS an IPO, so even better.

As for Cursor? Doesn't seem like a objectively bad acquisition, and with the upcoming IPO, the price isn't a big deal.
>>
>>108657552
seems like people who get in on the IPO at the start will be very rich. i am too lazy to learn how to do taxes to get in on it
>>
>>108656290
Congratulations and Condolences.
The probability of the AI bubble being a genuine bubble has been revised upward from 0.9 to 0.95.
The value of handwritten coding has been preserved.
pi-coding-agent just keeps getting more popular.
>>
Rockets?
Lemme guess they gave up on the plans to "conquer" mars.
>>
>>108656290
why is spacex training AI for coding again? 60b could build a whole fucking fab with the full supply chain that Musk wanted why would they pay for a company that has zero domain expertise.
>>
>>108658983
he can't deliver on those things, so he has to dump the money so he can keep saying that he doesn't have enough money to deliver on those things
>>
>>108657119
>SpaceX is massively overpaying because they need a big A.I. story before it goes public in June.
Threatening to pay $60B for Cursor isn't really 'a big A.I. story' tho. It's just another retarded acquisition at an even retarder price. Cursor spends 100%+ of their revenue on Anthro/OpenAI API costs, both of whom are probably also losing money on the deal their end. And throwing xAI into the mix will not be improving anything, for anyone.
SpaceX is now just this fantastic collection of shit that hemorrhages money or/and Musk overpaid for, or half-arsed plans for more moneypits like fabs. And this level of concentrated shit under one roof is not sustainable, even with an IPO.
>>
>30% more than Twitter which was already overpriced as fuck
>for a fucking coding wrapper which uses other companies models (cursor-small is nothingburger)

is this money laundering?
>>
>>108659417
It's the data more than anything. The prompts, the completions, etc.
>>
File: Untitled.jpg (110 KB, 498x498)
110 KB JPG
Kinda wild how Musk keeps consolidating garbage into SpaceX while trying to go public. What's his endgame here? Cash out when Trump's turn ends and move back to South Africa?
>>
>>108657650
That happened a while back when they announced the near term focus is data centers in space.
>>
>>108656290
>with SpaceX's million H1bs
>>
>>108659474
There's a certain scam logic to inflating an IPO to a size whereby index trackers etc have to buy the shit. Because that's exactly what they do. Musk is pushing *very hard* for the bypass of a whole bunch of rules that would normally prevent such an obvious pile of space junk and drug-fevered acquisitions being included in the main stock indexes at all, and they are currently bending over accommodatingly. tldr, he just needs to float 5% or so at some ludicrous valuation and the sheer scale of the shipwreck means he's automatically taking pension funds etc down with him. Except the Cap'n will have long fled the sinking ship.
>>
>>108657512 >>108659474
spacex was always used by elon to grift
the engineering and everything behind is serious, at least until tom mueller left, from there i am not so sure anymore, but he was always like that, he won the lottery when griffin got the nasa chair
>>108657650
it was always just an absurd pipedream, but he somewhat believed it for a while, i grant him that
that is how all this started, he contacting the mars society, griffin and the others, and financing them, they got him in in contact with tom mueller who was already developing the falcon engines. So he got a really good engineer and contacts of the next administrator from his mars plans, from there it was absurd claims after absurd claims and have people like the ars technica guy assliking him and validating every stupid made up bullshit as a 5d chess movement
we could also talk about how extremly ridiculous his plans to send a biological rover to mars were. but he actually never delivered what promised, yet people still belive their absurd price predicitons, even when they have changed them a ton in the past
>>
>>108656385
>>108656716
>>108659417

isn't cursor making billions of profit per year?
>>
File: G7cf6H1W4AAgxmq.jpg (878 KB, 3277x4096)
878 KB JPG
>>108657588
you don't. its illegal for you to get in on an IPO unless you are actually already massively (((rich.)))
>>
>>108656290
What did they buy exactly? Isn't Cursor just a VSCode fork with Kimi 2.5 with custom weights?

Also, Codex and Claude Code is at least one generation ahead of Cursor right now
>>
>>108660165
Yeah, it's a stupid move, the AI race in the west is a two company game. Musk's only good product was Grok's image and video capabilities, but with the recent paywall that shit will soon die too. I wonder if China is keeping up with OpenAI or Anthropic or if DeepSeek was just a fluke.
>>
>>108660269
>I wonder if China is keeping up with OpenAI or Anthropic or if DeepSeek was just a fluke.
I feel like China will always be behind, these AI Agent tools like Codex, Claude Code and Antigravity are the new IDEs, and China never built top tier IDEs or developer tools before
>>
>>108660116
It might be, if it didn't have to pay OpenAI and Anthropic. They do however and it doesn't (profit). They can just rely solely on xAI now. I'll leave you to guess where the potential quality issues inherent in that might be, the 'loss making service meets loss making service' aspect by side.
>>
What exactly is there left to code? The problem isn't lack of code; it's that existing code is slow as fuck.
>>
>>108660634
>It might be, if it didn't have to pay OpenAI and Anthropic.

aren't the users the ones paying?

https://e.vnexpress.net/news/tech/personalities/4-mit-graduates-who-built-the-popular-ai-coding-tool-cursor-become-billionaires-4965462.html
>>
Apparently they're paying for all data they have not for wrapper
>>
>>108660652
afaik, their outgoing API costs were running slightly North of 100% of revenue. For sure, they weren't losing more than most other AI Cos tho, and there doubtless is value there. When maybe not $60B worth.
>>
>>108660634
Personalized stuff I guess? It feels like anyone can start their little Photoshop like project or even a goddamn OS with AI helping. Open source Printer drivers would be sorta hilarious to make as well considering how fucking pozzed these are, HP would go full anudda shoaH.
>>
>>108660086
I gotta say, spaceship is starting to feel like Space X's cybertruck.
Something like Elon swaggering into the board room, drew on a whiteboard and said "make this, bitches"
>>
Sounds like xAI is betting that the IDE will remain the center of gravity for programming.
>>
>>108661633
reasoning is possibly more that Cursor were Anthropics biggest customer. But as they all lose money anyway, not sure how much of a loss they'd be.
>>
>>108656290

**Gets option to buy cursor for $60b**. Otherwise, they pay $10b for a new model from Cursor. These details matter.
>>
>>108661848
>Otherwise, they pay $10b for a new model from Cursor.
aka, chink open sauce brand, with a few tweaks?
nice work if you can get it.
>These details matter.
not to Musk they don't. Apparently.
>>
>>108656290
Buy an ad.
>>
>>108661286
Starship was supposed to be something like 4x as big but I guess that was too far even for sycophantic engineers so Elon settled for the little baby version. It was also supposed to be bare stainless steel but that doesn't work for re-entry (maybe they will do that for the ones that don't have to return though).
>>
>>108656290
in seriousness, my honest guess is that they're buying Cursor for their Composer AI because Grok, regardless how much Elon wants to make better, is still shit at code and somehow cursor team was able to ramp up a much more capable coding ai with fraction of cost and manpower
>>
>>108657479
>I just don't understand how business works any more
Cool it with the antisemetic remarks
>>
>>108660269
>I wonder if China is keeping up with OpenAI or Anthropic or if DeepSeek was just a fluke.
China is investing heavily in practical AI use cases in robotics and medicine.

Meanwhile here in the west we primarily use it for chatbots and vibecoded webslop.
>>
I feel like elon could've just made his own cursor equivalent company for much, much less than $60 billion
>>
>>108660634
idk but they probably said this in the 80s too
>>
>>108663412
It's weird how he can't just ask Grok to make it in 5 minutes. That should be well within its capabilities.
>>
>>108663412
thats because he already has an AI Co., so, he knows he can't. The shit he doesn't yet know he can't do either - chip fabs, yadda - he can happily go on spouting complete shite about how he, Elon, is gonna do it faster, bigger, better than everyone else. Until inevitably proven otherwise.
>>
>>108663541
post your rocket ship company eds-kun
>>
>>108663512
It's weird how Anthropic just rents out their AI to developers instead of using their AI to completely dominate the software industry and permanently shut down all competitors.
>>
>>108663558
I don't want one
personally, I think Mars looks a right shithole, but, each to their own.
>>
>>108659168
>And this level of concentrated shit under one roof is not sustainable
SpaceX holding a near-monopoly on space launches, Starlink being the ISP of choice for terrorists and campers and CIA assets alike, X holding insane soft power from being the place where all the politicians post publicly, Grok being considered a reliable source of anything for a concerningly large part of the population...
Like it or not but SpaceX as a company is not going anywhere anytime soon. Many of the things it holds are simply too big or too important to fail.
Of course it's a different question entirely whether Elon himself gets to stay on the ride till the end.
>>
>>108663569
It's weird how people never learn from seeing shovel salesmen standing along the road to riches.
>>
>>108656290
RIP
It was actually good. Almost singlehandedly caused vibe coding to be more than a short-lived meme.
>>
>>108663755
except in this case, The Big Shovel Co's sole product - A Big Fooken Shovel - is a loss leader.
>>
>>108656672
It is. But it's the Cursor that was very good at acting like a Genie.
>>
>>108657479
It's very simple.
60billen for owning Cursor, or 10billen to use whatever Cursor wants to cook with xAIs compute.
>>
>>108663797
>The Big Shovel Co's sole product - A Big Fooken Shovel - is a loss leader.
sure seems that way
>>
>>108663823
Seems like some kind of money laundering to me. There's nothing magic about Cursor - just make a legally distinct clone (or don't bother with the legal part since AI companies are above copyright law).
>>
>>108659168
>Cursor spends 100%+ of their revenue on Anthro/OpenAI API costs, both of whom are probably also losing money on the deal their end. And throwing xAI into the mix will not be improving anything, for anyone.
LMAO could be true. But one thing is certain. Cursor innovates. Everyone else steal.
I don't understand why they don't patent their stuff at Cursor. Are they fucking dumb? Apple and Microslop used to patent the most common sense things and were a bone in the throat of every Linux user for decades. These genious idiots at Cursor had a goldmine of well implemented and battle tested ideas and protected exactly 0 of 'em? What the fuck?
>>
>>108660086
>spacex was always used by elon to grift
>the engineering and everything behind is serious, at least until tom mueller left, from there i am not so sure anymore, but he was always like that, he won the lottery when griffin got the nasa chair
I don't think you know what that word means. If it was merely about milking the government for money why did't Boeing or Lockheed Martin develop reusable launch vehicles and put mega constellations in space. SpaceX undercut ULA launch prices by a lot, but somehow this is a grift.
>>
>>108663876
it's a grift because elon cut funding to his transsexual d&d club in sri lanka
>>
>>108660304
They have a chink IDE. But nobody cares. Afaik it says in ToS it will steal your shit right away.
>>108661823
For a reason. Sonnet used to be top 1 at coding when paired with Cursor.
>>108663836
> Seems like some kind of money laundering to me
Hard to argue, it sounds a bit inflated. Or not a bit, a lot. But I say what they say. Apparently they need compute and xAI wants to use whatever they are trying to produce.
> There's nothing magic about Cursor
It's good enough to be in magic tier. Literally everyone is trying to copy it and it's been like that for 3 years or so.
>>
>>108660605
You know that's not how it works, right? If you're providing access to a service belonging to someone else, you would have to be fucking retarded not to have some kind of markup. e.g. user pays $20 for access to ChatGPT, Claude, etc all in one place, but you only give them $10 worth of metered access to those LLMs. You make $10 profit. It's appealing to the user because they don't have to have a separate plan with OpenAI, Anthropic, etc and Cursor offers smaller plan options for people who don't want to pay $80/month for something they don't use a shitload of. I was a big Cursor user because I loved having access to all of the major models for a single price, until I discovered that I could host my own LLM on an old gaming laptop for free. It's slower than Claude, Grok, ChatGPT, etc but I don't have to pay a monthly fee and I don't have to deal with "I'm sorry, but that's against my programming, I can't do that" when I get too far into "black hat" territory, and having to hit it with the "hypothetically..." meme. It's worth putting up with the slowness from only having 6GB VRAM available to have it do what I fucking tell it to, and not have to pay for it.
>>
>>108664289
Anon 8GB VRAM was bare minimum years ago, what are you even doing?
Pro tip you never asked for: get yourself 32GB of RAM (4x 8GB is reasonably priced) and some 8GB GPU. Bare minimum is RX580, going any lower is unreasonable.
Grab any good MoE model, something like 27B gemma4. And get going. Not sure about cool hax0ring, but it's pretty uncensored in other areas.
>>
Grok is shit though
>>
File: tldr_it_loses_money.jpg (463 KB, 859x1540)
463 KB JPG
>>108664289
>you would have to be fucking retarded not to have some kind of markup
Yes.
and Cursor have been happily operating in the 'fucking retarded' zone since they started, with near every new customer cost them more than they brought in revenue. They have tightened their subscription levels some so they are losing slightly less money while attempting to make more use of their own proprietary model, Composer. But it all still loses money. As does virtually other AI Co.
>>
>>108661929
Jump into an acid vat
>>
>>108656290
wrong

$10B (the entire team)
or
$60B if they can prove they can develop a model better than opus (after the stock goes public and stock gets devalued)

spacex gets to decide, on their terms, they will absorb the entire team
>>
Why doesn't Elon just generate his own editor?
>>
>>108666137
They're training half a dozen other models. Musk is hedging all options at the same time to get the best outcome.
>>
>>108660116
I doubt that? Unless they did something revolutionary i doubt the API costs for their subscription would have made them anything but a loss.
>>
>>108656290
Even an AI would tell you this is a bad idea.
>>
>>108656290
>60 billion for a chatgpt wrapper
oh wait no they "made" (reskinned chinese kimi) their own model that one time
>>
File: 1776114305458597.jpg (66 KB, 940x940)
66 KB JPG
>>108656290
>>108656440
>SpaceX buys Cursor for $60 billion
>The world doesn’t make much sense anymore
I have no idea what these people are on anymore, they live in a completely detached reality of their own.
https://www.reuters.com/world/spacex-conquered-stars-now-eyes-bigger-opportunity-ai-2026-04-23/
>The AI unit posted an operating loss of $6.4 billion in 2025, sharply wider than the $1.6 billion loss a year earlier.
>Those losses eclipsed the $4.4 billion in operating profit generated by Starlink, SpaceX’s satellite internet business and its largest revenue engine, which brought in $11.4 billion of its $18.7 billion total revenue last year. Overall, SpaceX lost $4.9 billion.
>The S-1 regulatory filing, in which companies disclose their financials and key risks before going public, shows that SpaceX expects more than 90% of that market – or $26.5 trillion – could stem from the AI sector. The vast majority of that, $22.7 trillion, could come from AI for businesses.
>>108656672
It's the same chatbot as always, these people unironically think they will reach AGI with a chatbot and that will bring a post-scarcity interstellar civilization, they live in scifi world because they are not engineers/programmers, they don't know shit about technology
>>
File: seems_legit.png (27 KB, 705x366)
27 KB PNG
>>108668906
>Elon 'PT Barnum' Musk IPOs Present
that S-1 is an unholy conglomeration of wind, piss and more or less actual fraud.
>>
>>108666137
brand recognition matters
besides, AI copying other corporation's products is an anthropic thing
>>
File: Capture.png (276 KB, 762x498)
276 KB PNG
>>
>>108660116
no of course it doesn't, it's a silicon valley startup
kike liar
>>
>>108670543
>The parasites are mad they can't install retards who only care about the next quarter
Based
>>
thats more than twitter
>>
>>108677024
>The parasites
Uhm, that would be 'owners', legally speaking. No investation without representation.
>>
File: 1769105029713192.jpg (190 KB, 865x908)
190 KB JPG
>>108656775
>no one wants to work at the child porn factory.
speak for yourself
>>
>>108677494
>the child porn factory
Only place I'd willingly wageslave in.



[Advertise on 4chan]

Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.