Elliot (Dreams of Code) on AI, Rust, and the Future of Dev Education
Better Stackgo watch the original →
the gist
Elliot, creator of 'Dreams of Code', discusses why he is pivoting to Rust for 2026, the decline of traditional coding tutorials in the age of AI, and his pragmatic approach to using LLMs for scaffolding rather than architecture.
The Shift in Developer Education
Elliot observes a significant decline in the demand for traditional coding tutorials, noting that as agentic AI tools become more capable, the motivation for beginners to learn foundational syntax has plummeted. He argues that while AI can generate functional code for simple tasks, it struggles with non-trivial, scalable architecture. He maintains that deep, foundational knowledge remains essential for debugging and optimization, even if the current market sentiment favors quick AI-generated solutions over long-term skill acquisition.
The Rust Pivot and Systems Programming
Elliot is betting on Rust for 2026, viewing it as the primary language for high-performance, cross-platform development. He describes his current project—a native video editor—as a test of this thesis. He highlights the difficulty of managing cross-platform media pipelines (using AVFoundation on macOS, Media Foundation on Windows, and GStreamer on Linux) and notes that AI often fails to provide robust, modular solutions for these complex systems, instead offering brittle code that solves immediate symptoms rather than underlying architectural problems.
Pragmatic AI Workflow
Elliot advocates for a 'scaffolding' approach to AI. He uses LLMs to generate public interfaces, stubs, and boilerplate, but insists on writing the core logic himself. He has moved away from 'agentic' IDE experiences like Cursor, preferring the simplicity of Neovim. He finds that AI-integrated IDEs often clutter the interface with unwanted suggestions, whereas he prefers using CLI-based AI tools (like Codex) that allow him to maintain control over the code-reading and editing process without constant AI interference.
Tooling and Environment
Beyond AI, the conversation touches on the 'terminal renaissance.' Elliot emphasizes the importance of CLI tools like zoxide for productivity. He also discusses his preference for NixOS for reproducible environments, noting that while the learning curve is steep, the ability to manage dependencies and system configurations reliably is a significant advantage for a professional developer.