TypeScript, C# and Turbo Pascal with Anders Hejlsberg

2026-04-29 · Source: The Pragmatic Engineer · Field: Technology & Digital — Software Development & Engineering, Artificial Intelligence & Machine Learning · Depth: Advanced, extended

Summary

Anders Hejlsberg, creator of Turbo Pascal, Delphi, C#, and TypeScript, reflects on his 40-year career in programming language design and the future of software engineering with AI. He discusses the origins of Turbo Pascal as a fast, interactive, and affordable compiler-IDE package, and the evolution to Delphi for Windows GUI and client-server development. Hejlsberg recounts C#'s genesis from the Sun vs. Microsoft Java lawsuit and the need for a language combining C++'s power with Visual Basic's ease, featuring managed code, garbage collection, and a unified object system. He details TypeScript's emergence as an open-source superset of JavaScript, driven by the need for a type system to enable robust tooling, and its subsequent rise to popularity. The discussion also covers the design of async/await, the internal workings of the TypeScript compiler, and the increasing role of AI in code generation and review, emphasizing the importance of determinism and semantic understanding in AI-assisted development.

Key takeaway

For AI/ML engineers and software architects evaluating language choices or designing new systems, prioritize languages with robust type systems and strong locality. This approach not only enhances developer productivity through superior tooling but also improves the efficiency and reliability of AI-generated code by providing clearer semantic context and reducing token burn, ensuring deterministic behavior in critical applications.

Key insights

Effective programming language design prioritizes the entire developer experience, not just the compiler.

Principles

• Interactive development tools enhance productivity.
• Type systems are crucial for scalable tooling.
• Locality in code aids AI processing efficiency.

Method

Language design benefits from small, experienced teams critically evaluating ideas, followed by parallel specification writing and compiler implementation, evolving to self-hosting and interactive services.

In practice

• Adopt gradual typing for flexible development.
• Utilize AI for test generation and drudgery tasks.
• Focus on modular code to improve AI context handling.

Topics

• Programming Language Design
• TypeScript Development
• C# Language
• Integrated Development Environments
• AI in Software Engineering

Best for: Software Engineer, AI Engineer, Director of AI/ML

Related on AIssential

• AI Is Quietly Picking Your Next Programming Language For You — AI coding tools significantly favor statically typed languages like TypeScript due to enhanced contextual understanding.
• Build to Last — Chris Lattner talks with Jeremy Howard — Durable software requires deep understanding, robust architecture, and craftsmanship, resisting AI hype and superficial productivity metrics.
• I learned Odin — Rewriting a game in Odin offers a path to rediscover coding joy and explore a C-like language designed for game development.
• JavaScript Developers Will Be Playing With Fire In 2028 — Here is the Substantiation. — AI's rapid code generation capabilities challenge traditional JavaScript development roles, necessitating developer evolution.
• Bridging Functional Correctness and Runtime Efficiency Gaps in LLM-Based Code Translation — LLM-translated code often lacks runtime efficiency; SwiftTrans improves both correctness and speed through multi-perspective exploration and difference-aware selection.
• From Machine Code to Machine Intelligence: The Evolution of Programming Languages — Programming is evolving from line-by-line coding to directing AI systems that autonomously achieve outcomes.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by The Pragmatic Engineer.