Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

SpaceX to buy Cursor for $60B

reuters.com

Post reply on HN
Print view View original

991–1000 of 1001 posts

Re: SpaceX to buy Cursor for $60B

#991

Post by Ray20 »

danpalmer wrote: post #819
barredo wrote: post #25

>> SpaceX told investors during the IPO process that it sees an addressable market for AI products worth $26 trillion, roughly equivalent to U.S. GDP. I very much can imagine a future documentary in a few years. With the host asking the audience: "Where were the signs?"

AI hype is full of bad total addressable market hypotheses. There just isn't $26T of money to buy this stuff. Patrick Mackenzie, at the time at Stripe, about 10 years ago, talked about how Stripe's growth depended on growing the actual internet. Not just growing their market share within it. Google, Meta, etc, have projects dedicated to getting more people online, because when you have 2bn users, you kinda have to ma…

> There just isn't $26T of money to buy this stuff.

But there is. Like, it is less then GDP of US only

Re: SpaceX to buy Cursor for $60B

#992

Post by trees101 »

arglebarnacle wrote: post #721
kensai wrote: post #718

Earlier quoted context omitted.

"Prompt "engineering" and working around your tools will be over in a year or so." What do you mean by that? What is happening in just over a year or so?

The models are getting better at agentic coding, so over time using complicated harnesses and precise prompt engineering to attempt to squeeze out an extra X% performance will become irrelevant as the models approach expert-level performance. The bitter lesson in miniature.

there will always be a difference between the general capabilities, and the particularities of your exact environment and requirements.

Closing this gap is done in the harness, either through Skills, user behaviour/prompts , Agents.md etc etc.

I think that this is an area worth investing time in, but it is indeed hard to know what the scope of this is.

Re: SpaceX to buy Cursor for $60B

#993

Post by slashdev »

danpalmer wrote: post #819
barredo wrote: post #25

>> SpaceX told investors during the IPO process that it sees an addressable market for AI products worth $26 trillion, roughly equivalent to U.S. GDP. I very much can imagine a future documentary in a few years. With the host asking the audience: "Where were the signs?"

AI hype is full of bad total addressable market hypotheses. There just isn't $26T of money to buy this stuff. Patrick Mackenzie, at the time at Stripe, about 10 years ago, talked about how Stripe's growth depended on growing the actual internet. Not just growing their market share within it. Google, Meta, etc, have projects dedicated to getting more people online, because when you have 2bn users, you kinda have to ma…

If that were to become reality, and I won’t guess at the odds of that, is because it grew the world economic pie. Thinking about it in terms of the size of the pie today is treating it like a zero sum game. It’s a different game.

Re: SpaceX to buy Cursor for $60B

#994

Post by anon7000 »

estetlinus wrote: post #466
frays wrote: post #102

Is anyone on HN still actually using Cursor in 2026? Everyone I've spoken with is now using either Claude Code or Codex (or Copilot because their companies force them to).

I have seen a few codebases lately with AI-bullish teams. Code produced by Cursor reeks of low quality. I’ve tried it but never got hooked. AFAIK their market is pseudo-technical people who haven’t found the terminal yet.

I’m running Opus through Cursor like I would in Claude code, so make sure you’re not comparing different models.

But the core difference for me is I can easily see the code changing in cursor, but not in Claude code. Bigger tasks, I’ll have to drop into other tools to see what Claude is doing along the way. I don’t like that. I like everything being in the same spot

Re: SpaceX to buy Cursor for $60B

#997

Post by zs234465234165 »

marcuschong wrote: post #804
01100011 wrote: post #421

I stopped using Cursor when I started getting comfortable with Codex/Claude. Cursor is just annoying with the constant popups and it's just not as good. Now my workflow is to use my normal editor, add a todo describing what I want, and then ask Codex+gpt-5.5 to implement it. It absolutely nails it. Using codex is so much more like working with a partner vs the noise and annoyance of Cursor. That said, I think we're i…

It's curious that the person claiming LLMs will soon skip code entirely and go straight to binary is willing to spend $60bn on Cursor.

I’m sure he has a good reason

Re: SpaceX to buy Cursor for $60B

#998

Post by coffee »

jr3592 wrote: post #578
sergiotapia wrote: post #500

Earlier quoted context omitted.

On the flipside, I enjoy Cursor now and came back to it after leaving it over a year ago. The 2.5 model is fast as hell and very good. And whatever harness they have it's terrific, great results. I also really enjoy the fact that I can open my website in the Cursor in-app browser and just click and reference stuff. It's a really cracked workflow. The models can only get better for them.

I would also add that Cursor's "Debug" harness is incredible. Hit "Tab" in the AI editor to Tab through the options (Plan, Multitask, Ask, etc.) If you do any kind of on-device work, it will spin up a local HTTP log server, and pipe logs from your real device (phone, hardware, etc.) to the server and do realtime debugging. Claude will mostly guess, have you copy + paste logs, etc.

this

Re: SpaceX to buy Cursor for $60B

#999

Post by z3t4 »

kopirgan wrote: post #851
ghshephard wrote: post #626

Earlier quoted context omitted.

I use cursor 8+ hours/day at work, and have full (and effectively unlimited) access to Claude Code and Codex - tools which I also use personally. I suspect that your "constant popups" were when you were using the editor - a mode that I'll confess I haven't touched in 3+ months. Workflow in Cursor is actually awesome - I'm a little outdated in how I use it - I still establish goals/objectives, rather than managing the…

Not trying to be funny but seriously, if these tools can produce a tested 'product' in 45m, shouldn't we be seeing millions of them out there? I mean how far are we from a fully AI built Oracle ERP or even a notepad or helix?

Have you ever heard "I can do that in a weekend" and they usually can. The difficult part is not building the product, it's selling and marketing, the buisness part. It's quite common buisness tactic to outright copy someone else's product or buisness.

Re: SpaceX to buy Cursor for $60B

#1000

Post by z3t4 »

jw1224 wrote: post #491
devin wrote: post #475

Earlier quoted context omitted.

What are you saying is going to be over in a year or so?

Not OP, but I generally agree. Models are powerful enough now to reliably instruct other models. They don’t need fancy tools or IDEs, just the command line. With deterministic workflows, type-safe languages and test suites, agentic loops pretty much “can’t fail”. They will continue until the types resolve, the tests pass, and the project requirements are deterministically met. By that point it’s literally just a case…

It works great in dynamic languages as well. Static typing is mostly to aid the IDE. In dynamic languages you can infer the type by looking at the code, and LLM are good at that.
Post reply on HN