compiler

Definition

Plain language

The program that turns the code a developer writes into something a computer can actually run.

Also called: compiles, compile, compiled

Why it matters: It is the gatekeeper that turns written code into a runnable program and refuses anything that breaks the language's rules.

For example, when a developer hits 'build,' the compiler turns their human-readable code into the instructions the computer actually executes, flagging any errors along the way.

Heard on the show

“The program compiles or it doesn't.”

Mentioned in 25 episodes

1. 240
   Frontier Models Designed Follow-Ups To Fraudulent Papers 93% Of The Time
2. 203
   The Thought a Model Doesn't Say — and the Lens That Reads It
3. 188
   A Coding Agent Found a Hole in a Peer-Reviewed STOC Proof for Five Dollars
4. 181
   How to Backpropagate Blame Through a Team of Chatbots — And When It Backfires
5. 177
   Why Raw Profiler Data Made an AI Worse at Writing GPU Code
6. 176
   An AI Designed Its Own Psychology Studies, Then Confirmed What It Found
7. 139
   When Optimizing One GPU Kernel Quietly Breaks the Whole System
8. 132
   The Agent Failed — But Did the Instructions Deserve to Be Followed?
9. 131
   Why Autonomous Research Agents Forget Their Own Lessons, and Arbor's Fix
10. 125
    AI Coding Agents Run a Marathon, and Fewer Than One in Three Finish
11. 122
    When Your Coding Agent Lies About the Fix: Verifying the Plan Before the Model Runs
12. 121
    When the Agent Says It's Done But Nothing Happened: Debugging the Harness, Not the Model
13. 117
    How an Open AI System Verified 672 Hard Math Proofs for Under $300
14. 110
    How an Agent Got 44 Points Better by Mining Its Own Scratch Paper
15. 101
    Treating Math Formalization Like a Codebase, and Where the Agents Cheat
16. 088
    Two Levers for Self-Improving AI: When Rewriting Code Isn't Enough
17. 079
    An Old Idea From Cognitive Psychology Reshapes How We Reward Reasoning Models
18. 078
    Training a Markdown File: When LLM Self-Improvement Borrows the Discipline of Neural Net Training
19. 075
    Growing Code and Proof Together: Verified Systems in Ten Hours Instead of a Year
20. 067
    An AI Just Solved a 1996 Erdős Problem—and the Simplest Agent Won
21. 063
    Why Web Agents Are Slow: A Compiler-Style Fix for Computer-Use Latency
22. 034
    Catching Multi-Agent Deadlocks Before Deployment With a 40-Year-Old Tool
23. 027
    When AI Agents Build the Serving Stack: A Bet on Bespoke Infrastructure
24. 024
    An AI Agent That Found 28 Zero-Days in Windows — And What Made It Work
25. 014
    Why a Constrained Pipeline Beat a Full Coding Agent at Finding Bugs 30-to-1

Related concepts

Related terms