HomeArticle

SpaceX acquires Cursor, Musk dives deep into coding: the new focus of the AI race is set

雷科技2026-06-18 08:26
AI Coding is becoming increasingly important.

To be honest, Lei Technology (ID: leitech) really didn't expect that SpaceX's first major deal after going public would be to buy Cursor.

Although Musk revealed this agreement to the public as early as April, spending $60 billion is still an unimaginable big move for many people. Of course, if you really dig deeper, a lot of this is SpaceX stock in a high - premium state, and the actual cash Musk took out is far less exaggerated than this figure.

But this doesn't prevent Cursor from becoming one of the "miracles" in business history.

After all, the valuation inflation rate of Cursor even exceeds that of Google and Facebook in the Internet era. How can an AI Coding company that has been established for only a few years get such an exaggerated price? Lei's answer is: It's all about AI Coding. More precisely, it's about the position of AI Coding in the future AI system.

Image source: Cursor

Today's "Coding" is no longer just programmers sitting in front of a computer typing code. It is becoming the underlying ability for AI to understand the world, transform systems, call tools, and complete tasks. In other words, in the future, all AI Agents have to go through Coding to really do their jobs.

In this sense, when SpaceX bought Cursor, it wasn't buying a developer tool, but an entrance to the future AI execution layer.

SpaceX Buys Cursor, Musk's "Big Make - up" on Coding

What's so special about Cursor? Well, from an industry perspective, Cursor is the first code editor that is truly "AI - native", rather than simply adding an "AI plugin" to an existing code editor.

Based on this feature, Cursor allows developers to understand, create, and modify complex code libraries across multiple files through pure natural - language conversations. Lei once saw someone on the foreign network describe the difference between Cursor and ordinary AI programming tools like this: The latest Tesla intelligent driving compared to the adaptive cruise of an old - fashioned fuel - powered vehicle.

This is also why Cursor has become one of the fastest - growing SaaS companies in history. Starting with $100 million in revenue at the beginning of 2025, it only took one and a half years to grow to an annualized revenue in the billions of dollars. Behind this are tens of thousands of enterprise engineering teams and more than 60% of the world's top 500 companies.

So the real value of Cursor doesn't lie in how many individual developer users it has, nor in how much subscription fees it can collect each month. It's that it is becoming the "software productivity" of many enterprises.

Image source: techstartups

Although NVIDIA has become one of the world's highest - valued companies with its hardware, there is an undeniable fact: software is the core framework of the digital world. From enterprise management to manufacturing, you'll find that almost everything depends on software. And as an individual in society, "I" am also becoming more and more dependent on software in aspects such as food, clothing, housing, and transportation.

This also means that as long as AI wants to enter real - world scenarios, it must be able to understand, generate, and modify software systems.

And this is the significance of Cursor to SpaceX. Although SpaceX is ostensibly a space company, it actually represents Musk's AI territory. SpaceX has engineering systems and computing power infrastructure, xAI has models, X has data and distribution channels, Tesla has autonomous driving and robots, and Starlink has a communication network.

Then you'll find that Musk has almost everything, except for an entrance that can really enter the software production site. So this deal is actually Musk filling in the last key piece of the puzzle for his AI ecosystem.

Cursor is not the core; Coding is the key.

AI Coding Is Not Just about Software, but also a Computing Power Business

Although Cursor is important, is its $60 - billion valuation too high? If we only look at it from the perspective of a developer tool, the expected future revenue is really hard to support this price (even considering Musk's strategic needs).

But the real focus is that the AI Coding represented by Cursor is not a simple "software business", but an intersection point between future AI computing power, model capabilities, and the software foundation.

In the past year, when people talked about AI, the most frequently mentioned thing was the "entrance". You'll find that many software UI interfaces have an "AI" button, all wanting to be the first entrance for users to interact with AI. However, these entrances are mostly for ordinary users and solve relatively daily problems.

AI Coding is different. It faces the software system itself. Code is not only the underlying language of the digital world, but also the key for AI to move from "being able to talk" to "being able to do things".

Image source: Lei Technology

This is why Lei believes that Coding is not just for developers. In the future, Coding will be the foundation for all AI. For example, when you ask AI to help you book a flight ticket, make a form, or formulate a plan, it seems like you're having a conversation with a chatbot, but what it really has to do in the background is to transform your intention into a series of executable logics, tool calls, and system operations.

This process is essentially "Coding" in a broad sense. It's just that in the past, it was about coding code, and now it's about "Coding" for your needs. In other words, in the future, Cursor can actually transform into a kind of "AI operating system". After all, there isn't much essential difference between cross - file environment programming and cross - application operations.

OpenAI is actually doing something similar. The Codex ability of ChatGPT, although initially more focused on developer scenarios, has now gradually become a product of the integration of Coding ability and ChatGPT. ChatGPT can run code, process files, generate forms, and analyze data. Codex can perform tasks in the code library, and finally, workspace agents further embed this ability into the workflow.

Image source: codex

From this perspective, AI Coding can no longer be regarded as a branch of AI applications. Instead, it may become the foundation for all future complex AI applications.

This is also the reason why computing power giants must compete for AI Coding. Because the deeper AI Coding goes, the more computing power it consumes. Then, whoever controls the AI Coding entrance can control the model - calling scenarios of enterprises and users, and then create more computing power demand.

Buying and selling to oneself, this is the most profitable business in the world.

Where Will the No - Longer "Independent" Cursor Go?

After Cursor was acquired by SpaceX, the most concerned question among developers is whether it will still be as open as before.

One of Cursor's greatest past advantages was its neutrality. It is not the official client of a certain model company, but a relatively open AI Coding workbench. Developers can connect to different models according to their needs and build workflows around their project habits.

This openness is very important because developers don't like to be locked in a fixed ecosystem. Especially AI developers, they are used to using the best one. They use Claude today, GPT tomorrow, and try Gemini or GLM the day after tomorrow. This is completely different from the situation in the past where they were not willing to switch once they got used to a certain programming language.

Moreover, there is a very sensitive question: How will it handle the data and code encountered in work? It used to make big enterprises feel at ease when it was an independent company, but this time the acquirer is SpaceX, which has a wide range of businesses. Enterprises that have business conflicts with SpaceX may have to reconsider their usage status.

People are also curious about whether Cursor will be more deeply bound to Grok in the future after being incorporated into the SpaceX system, and whether the third - party access policy will change. Such concerns are not unfounded. A recent example is OpenClaw. After its founder joined OpenAI, Anthropic announced restrictions on OpenClaw users' random calls to Claude's API, and Google also once restricted the use of some OpenClaw users.

Image source: OpenClaw

So, enterprises are worried about compliance issues, while individual developers are worried about whether Cursor is still easy to use. Everyone has their own troubles.

In addition, from the perspective of domestic enterprises, the acquisition of Cursor by SpaceX is actually a reminder: AI Coding is not a lightweight ability that can be easily outsourced to overseas tools. As software that can access code libraries, R & D processes, and enterprise core assets, once it is controlled by an overseas giant, enterprises must more carefully consider security risks.

This also brings new opportunities to domestic AI enterprises such as Zhipu. From CodeGeeX to the GLM series of models, Zhipu has been continuously investing in code generation, multi - language programming, agents, and developer tools. The newly released GLM - 5.2 already has various capabilities such as "end - to - end full - link development" and "deep system debugging".

Image source: Zhipu

However, Lei believes that for Zhipu, it doesn't necessarily have to create a "Chinese - version Cursor". Instead, it should establish an AI Coding system more suitable for the local market around the real needs of Chinese enterprises.

For example, government and enterprise customers are concerned about privatization and permission management, financial customers are concerned about auditing and compliance, and manufacturing customers often have a large number of old systems, industry codes, and internal processes that need to be integrated. Whether from the perspectives of localization, stability, or security, these needs cannot be easily met by overseas general tools.

So the opportunity for domestic AI Coding enterprises doesn't lie in simply replicating the interface or capabilities of Cursor. It lies in integrating a series of needs such as models, tools, and security into a complete AI Coding ecosystem.

Coding Has Become the Core Battlefield of AI

With the acquisition of Cursor by SpaceX, in the next step, the AI battlefield may shift to a more fundamental level.

For AI to move from "being able to chat" to "being able to do things", it must enter real - world scenarios. It needs to transform a person's words into a series of executable actions, and Coding is the bridge connecting natural language and digital actions. It can be said that Coding is no longer just a developer's skill. It has become the basic ability for AI to understand and transform the digital world.

For Chinese AI enterprises, the most important thing to realize is not to regard AI Coding as a small track, nor just to create a "programming plugin". Instead, it should be regarded as the hub between AI Agents, enterprise agents, the implementation of domestic models, private deployment, and the commercialization of computing power.

Today, Cursor seems to be a code editor, but tomorrow, AI Coding may be the operating platform of the AI system.

In the past, we said that software is devouring the world.

Now, maybe we should change the statement: AI is rewriting the world through Coding.