Developer Matt Webb Used ChatGPT to Learn Quaternion Maths

👁 117 reads

App developer Matt Webb has shared an illuminating example of how AI chatbots can serve as effective learning tools for complex technical subjects. In a post dated 21 August 2026, Webb explained that he successfully used ChatGPT as an interactive tutor to understand quaternion mathematics well enough to ship the augmented reality mode in his app, Galactic Compass 2. This achievement came after traditional learning methods, including books and conversations with mathematician friends, had failed to give him the understanding he needed.

How ChatGPT Helped Webb Learn Quaternion Mathematics

Quaternions are a mathematical system used to represent rotations in three-dimensional space, making them essential for augmented reality applications where virtual objects need to rotate smoothly and accurately as users move their devices. For many developers, quaternions represent a significant learning hurdle because they work differently from the more intuitive Cartesian coordinate systems most programmers learn first.

What made Webb’s approach noteworthy is that he did not ask ChatGPT to simply write the rotation code for him. Instead, he used the AI tool to build foundational mathematical understanding, then implemented the feature himself. This distinction matters because it demonstrates active learning rather than passive copying. Webb could ask follow-up questions, request clarifications, and work through concepts at his own pace until the mathematics made sense in the context of his specific application needs.

AI Tools as Patient Interactive Tutors

Webb described his experience with ChatGPT as having access to a patient, interactive tutor. This characteristic addresses a common challenge in technical learning: the need for immediate, judgment-free feedback when working through difficult concepts. Unlike a textbook, which presents information in a fixed sequence, or a human expert, who may not always be available or may inadvertently make a learner feel inadequate, an AI chatbot can answer the same question multiple times in different ways until understanding clicks.

As quoted by Simon Willison on his weblog, Webb stated: “With a patient, interactive tutor, I was able to finally do what I hadn’t by reading books and asking mathematician friends - I learnt how to use quaternions just enough to make the app work.” This practical, goal-oriented learning represents a pragmatic approach where the developer gained sufficient understanding to solve his specific problem rather than pursuing comprehensive theoretical mastery.

What This Means for Developers and AI-Assisted Learning

Webb’s conclusion directly challenges a widespread concern that AI tools might reduce intellectual curiosity or create dependency that weakens problem-solving skills. Instead, he found the opposite effect. “So learning doesn’t stop just because I outsource a bunch of thinking to AI,” Webb wrote. “It pushes me to learn more.” This observation suggests that when used appropriately, AI tools can lower the activation energy required to engage with intimidating subjects, ultimately enabling rather than replacing human learning and creativity.

For developers working on augmented reality applications, computer graphics, robotics, or game development, this example illustrates a practical workflow: use AI chatbots to build conceptual understanding of complex mathematics, then apply that knowledge to write your own implementation. The result is both a working feature and genuine learning that persists beyond the immediate project.

Frequently Asked Questions

Did Matt Webb have ChatGPT write his code?

No. Webb was explicit that he did not ask the model to write the rotation code for him. Instead, he used ChatGPT to build mathematical understanding, then implemented the augmented reality feature himself.

What app was Webb developing?

Webb was working on Galactic Compass 2, specifically implementing an augmented reality mode that required understanding quaternion mathematics for handling rotations.

Did using AI reduce Webb’s learning or curiosity?

According to Webb’s own account, the opposite occurred. He stated that learning does not stop just because he outsources thinking to AI, and that it actually pushes him to learn more, contradicting concerns that AI tools dull curiosity.

// read moreHow Many ChatGPT Conversations Are Used for Training? The MathGuidesAI Developers Reconsider as Claude Fable’s Cost Outweighs GainsNewsGitHub Models shuts down, leaving developers to find LLM accessAI Tools