

Episode 252: Sean Parent on Rust and AI
10 snips Sep 19, 2025
Sean Parent, a senior principal scientist at Adobe, dives into the world of Rust and AI. He shares insights from his work on a Rust-based no-code/low-code project, discussing the advantages of static typing and Rust's tooling. Sean also addresses challenges like verbose syntax and borrow-checker ergonomics. Furthermore, he explores the role of AI in coding, highlighting tools like Cursor and GPT as valuable aids for learning, while cautioning against the pitfalls of AI-generated outputs.
AI Snips
Chapters
Books
Transcript
Episode notes
Bootstrap Guard Rules For Repeating AI Errors
- If an AI repeatedly makes the same mistake, prompt it to generate a self-rule or guard you can paste into its context to prevent recurrence.
- Use the model's ability to craft constraints for itself and then apply them as explicit system prompts.
Property Models Enable Intent-Driven UIs
- Property-model libraries let you declare relationships between properties and have the system solve constraints under the hood.
- That approach makes UIs and intent-driven systems more compact and provable compared with imperative code.
Transitioning A Legacy System To Rust
- Sean Parent is writing production Rust for a new Adobe project after decades in C++ and prior implementations in C++ were long-lived at Adobe.
- He moved to Rust for better static typing, introspection, and error reporting while rewriting a property-model system.