Cursor’s CEO explains why AI coding still needs humans in the loop

Michael Truell on AI coding

Michael Truell, cofounder and CEO of Cursor, is changing how programmers think about AI coding. At a recent Fortune Brainstorm AI conference, he explained that programming no longer means typing line by line into a blank screen. AI now allows coders to step back while still building complex projects.

According to Truell, AI can handle end-to-end coding tasks, but programmers still need to know what’s going on under the hood. This balance helps prevent mistakes while letting coders leverage AI efficiency. His perspective reflects a modern shift in programming culture.

Vibe coding is not recommended

Truell cautions against the so-called “vibe coding” method, where users let AI handle everything without checking the code. He likened it to building a house without inspecting the floorboards or wiring. It may be fun for quick mockups, but risky for serious programming projects.

Vibe coding can work for games or small websites, but larger or advanced projects may collapse if left unchecked. Truell warns that blindly stacking AI-generated code can lead to shaky foundations, making problems multiply as the project grows.

Cursor embeds AI in coding

Cursor integrates AI directly into the development environment where programmers write their code. By understanding the context of existing code or even a whole codebase, it predicts next lines, generates full functions, and helps debug errors efficiently.

This approach allows developers to control the details while still benefiting from AI assistance. Cursor offers features like multi-line autocomplete, full function generation, and error explanations, bridging the gap between human expertise and AI speed.

From MIT project to billion-dollar startup

Truell and three fellow MIT graduates began the project in 2022 and launched Cursor publicly the following year. What began as a small project quickly grew into a popular coding assistant. Its growth reflects a shift in how developers approach coding.

LinkedIn lists Cursor with between 201 and 500 employees and public reporting placed annual recurring revenue in the low hundreds of millions by spring 2025, a sign of fast growth though not the one billion dollar figure sometimes claimed.

Cursor’s funding milestones

Cursor raised an $8 million seed round led by OpenAI’s Startup Fund in October 2023. Subsequent rounds attracted top Silicon Valley investors, including Andreessen Horowitz. These investments helped the company scale rapidly and enhance its AI capabilities.

In November 2025, Cursor announced a Series D of $2.3 billion at a $29.3 billion post-money valuation. The trajectory shows strong investor confidence and highlights the growing importance of AI coding tools in the software development industry.

Cursor helps coders focus

Unlike vibe coding, Cursor lets programmers dive into the details when needed. Users can control the AI to write specific functions or debug complex code while still using AI to handle repetitive or end-to-end tasks. This combination improves efficiency without sacrificing quality.

Truell emphasizes that this dual approach gives coders the best of both worlds. They can experiment and accelerate development while keeping a close eye on critical elements, reducing the risk of errors that might occur with fully hands-off AI coding.

Multi-line AI autocomplete explained

Cursor’s AI predicts the next line or multiple lines of code based on context. This helps developers write faster and reduces repetitive tasks. Instead of coding every detail manually, programmers can rely on AI to streamline routine workflows.

The feature also aids learning. Coders can see suggested implementations, debug approaches, and explanations of errors. It’s a way to blend hands-on coding with AI guidance, giving both new and experienced programmers a practical advantage.

Full function generation feature

Cursor can generate entire functions automatically, saving time and helping developers prototype quickly. The AI considers existing code, variable context, and user intent, allowing coders to focus on logic and higher-level structure instead of tedious implementation details.

This empowers programmers to innovate without getting bogged down in repetitive tasks. It also reduces common coding errors, making development faster and more reliable, especially for large or collaborative projects.

Debugging and AI explanations

Cursor assists developers by explaining code errors and suggesting fixes. This is particularly useful for new programmers or those unfamiliar with complex codebases. The AI breaks down problems step by step, making debugging more intuitive.

By combining explanation with automated code generation, Cursor helps developers maintain control while increasing productivity. This reduces frustration and accelerates project timelines, proving that AI can complement human expertise effectively.

AI coding still needs human judgment

AI coding assistants like Cursor are redefining how developers work. These tools allow programmers to innovate faster while maintaining code quality. But still, AI coding assistants rely on human judgment to review logic, catch errors, and guide design.

Truell’s insights highlight the shift from manual coding to AI-assisted workflows. The landscape of programming is evolving rapidly. Developers must supervise outputs, fix edge cases, and make final decisions that machines cannot reliably handle alone in production systems.

Cursor reaches one million users

By early 2025, Cursor had grown to one million daily users. Its combination of AI-assisted coding, multi-line autocomplete, and full function generation attracted developers from around the world. The platform has quickly become a go-to tool for modern programming projects.

The user growth highlights how coders are embracing AI for both speed and accuracy. Cursor’s integration into the development workflow makes it easy for both novice and experienced programmers to take advantage of AI assistance without sacrificing control.

Truell on coding balance

Michael Truell emphasizes that AI coding should balance automation with hands-on oversight. While AI can handle repetitive or end-to-end tasks, developers must still review and guide critical portions of code. This prevents mistakes and maintains the integrity of complex projects.

This approach helps programmers learn, innovate, and maintain high-quality code. Truell believes that understanding the details while leveraging AI efficiency is the key to successful modern coding practices.

Want to see how Claude is pushing the boundaries of creativity? Dive into the race with Gemini and Twitch in Pokémon Red.

The future of AI coding

AI coding tools like Cursor are changing how developers work. By combining deep control with automated assistance, coders can innovate faster without sacrificing accuracy. Truell’s perspective shows the shift from manual programming to intelligent, AI-assisted workflows.

The programming landscape is evolving rapidly. Tools like Cursor let developers use AI effectively while staying in control.

What happens when a startup fully commits to vibe coding? See how one Swedish AI company used it to smash every growth record.

What do you think about AI-assisted coding? Share your thoughts.

This slideshow was made with AI assistance and human editing.

Don’t forget to follow us for more exclusive content on MSN.

Read More From This Brand:

• a16z finds startups spending on vibe coding
• The creator of Claude code breaks down where vibe coding falls short
• Using AI coding assistants may reduce speed for veteran developers