How did Michael Truell get into programming?

Michael Truell first encountered programming when trying to create a mobile game with his brother, diving into Objective C despite its initial complexity. This sparked a long-term interest that led him to pursue programming further.

What were some of Michael Truell's early AI projects?

Before Cursor, Michael Truell worked on projects like a robotic dog that could be trained with feedback, exploring genetic algorithms, neural networks, and reinforcement learning. He even implemented his own small neural network library for microcontrollers due to memory constraints.

What was Cursor's first product idea before focusing on code completion?

Cursor's initial idea was to build an AI co-pilot for mechanical engineers, aiming to predict actions within CAD software like SolidWorks and Fusion 360. This project involved extensive data scraping and infrastructure development.

Why did Cursor pivot to focus on AI in coding?

After early projects like the CAD co-pilot and an encrypted messaging system failed to gain traction, the founders realized their core excitement lay in the future of coding. Seeing the potential for AI to fundamentally change software development, they decided to focus on building the best way to code with AI, despite the competition like GitHub Copilot.

How did Cursor develop its first editor?

Cursor initially built its own code editor from scratch using open-source building blocks over three months. They quickly learned that building a feature-complete editor was a massive undertaking and shifted to basing their product on VS Code to focus on AI features.

What drove Cursor's rapid growth in 2024?

Cursor's growth was driven by continuous product improvements in AI capabilities, such as codebase awareness, predictive actions, and faster AI responses. The product's effectiveness led to strong word-of-mouth adoption, especially among top builders in the YC community.

What advice does Michael Truell have for aspiring engineers and founders?

Truell advises focusing on working with people you respect and enjoy, and dedicating serious effort to projects you are genuinely interested in. He emphasizes building something over time rather than just checking boxes, and that foundational skills like programming and math remain valuable.

Michael Truell: Building Cursor At 23, Taking On GitHub Copilot & Advice To Engineering Students

Y Combinator

Science & Technology3 min read28 min video

Sep 3, 2025|54,815 views|1,152|47

YC Y Combinator

Save to Pod

Key Moments

TL;DR

Michael Truell built Cursor, an AI coding tool, at 24, achieving $100M ARR. He advises students on passion and perseverance.

Key Insights

Early programming interest and entrepreneurial drive, stemming from childhood game development, shaped Truell's path.

Persistent exploration of AI, from robotic dogs to data-efficient reinforcement learning, laid the groundwork for Cursor.

Despite initial failures with unrelated startup ideas, the team's conviction in AI's future in coding led to Cursor's creation.

Cursor began as a custom-built editor, but quickly pivoted to a VS Code base to focus on core AI features.

Growth was driven by product improvement, word-of-mouth, and aligning with the evolving needs of developers.

Truell emphasizes that fundamental programming skills remain crucial, even as AI transforms software development.

THE SPARK OF INNOVATION: EARLY BEGINNINGS AND AI FASCINATION

Michael Truell's entrepreneurial journey began in middle school, driven by a desire to create commercial ventures, starting with a foray into mobile game development. Despite an initial encounter with the complexities of Objective-C, his persistence led him to explore programming further. Simultaneously, his interest in AI sparked early on, leading to collaborative projects like building a robotic dog that could be taught through feedback, akin to training a pet. This early exploration into reinforcement learning and neural networks, even resulting in the creation of custom ML libraries for resource-constrained microcontrollers, laid a foundational understanding of AI principles.

PIVOTS AND PERSEVERANCE: THE ITERATIVE PATH TO CURSOR

Truell and his co-founders initially pursued various startup ideas after graduating from MIT, including a co-pilot for mechanical engineers requiring extensive data scraping and model training. They also explored an end-to-end encrypted messaging system. Despite significant effort and technical achievement, these projects failed to gain traction due to market fit issues or inherent technical trade-offs. This period of iteration and learning from un successful ventures was critical in shaping their understanding of product development and market viability, underscoring the importance of genuine excitement for the problem domain.

THE VISION FOR CODING'S FUTURE: BETTING ON AI TRANSFORMATION

The genesis of Cursor emerged from a profound belief that the entire landscape of coding was poised for a significant transformation driven by AI over the next five years. While tools like GitHub Copilot were gaining traction, Truell and his team felt that many existing efforts were merely incremental improvements rather than aiming for a fundamental shift in how software is built. They recognized a gap in the market for a product that truly embraced the potential of AI to automate coding tasks and redefine software development workflows, feeling a strong conviction to pursue this ambitious vision.

FROM SCRATCH TO VS CODE: BUILDING THE CURSOR EDITOR

The initial development of Cursor involved building a custom code editor from the ground up, incorporating various open-source components. This ambitious undertaking allowed them to integrate their core AI features, such as a universal remote for AI commands and early code suggestions. However, after realizing the immense undertaking of creating a feature-complete editor comparable to established tools like VS Code, they strategically pivoted to basing Cursor on the VS Code ecosystem. This allowed them to leverage existing infrastructure and focus their resources on advancing their unique AI capabilities.

ACCELERATED GROWTH AND WORD-OF-MOUTH ADOPTION

Cursor experienced explosive growth in 2024, scaling from $1 million to $100 million ARR. This was largely attributed to continuous product improvements, particularly in codebase awareness, predictive actions, and faster AI model integration. The market's strong emphasis on user preference meant that as Cursor became demonstrably better, word spread rapidly. Early evangelism through social media, especially by one co-founder's insightful posts on AI developments, and compelling demos also played a role. However, sustained growth was primarily fueled by organic word-of-mouth referrals among developers.

THE ENDURING VALUE OF PROGRAMMING SKILLS IN AN AI ERA

Looking ahead, Truell stresses that programming, much like mathematics, remains a fundamental and valuable skill, regardless of AI's advancements in coding. He highlights that the practical skills acquired through computer science studies foster crucial learning abilities that transcend specific technologies. For aspiring engineers and students, his advice is to focus on working on intrinsically interesting projects with respected peers, emphasizing deep engagement and building over merely checking boxes. This approach cultivates resilience, innovation, and adaptability in the face of rapidly evolving technological landscapes.

Mentioned in This Episode

●Software & Apps

●Organizations

●Studies Cited

●Concepts

Common Questions

Michael Truell was inspired by the rapid advancements in AI and the belief that all of coding would change in the next five years. He saw an opportunity to build a product that took this transformation seriously, unlike existing offerings that were only making incremental improvements.