
LLM as a Judge: Why Your AI Might Be Marking Its Own Homework
From Coding Chats by John Crickett
April 30, 2026 · 1h 7m
About this episode
John talks to Laura Dietz about the implications of using AI for evaluation in software development.
Coding Chats episode 76 - John talks to Laura Dietz - a computer science professor whose work focuses on whether AI evaluation metrics actually tell the truth. She's known for her critical take on "LLM as a judge" — not because she thinks it's useless, but because she wants numbers that mean something rather than numbers that just make a system look good. The conversation tackles some uncomfortable realities for software engineers: using an LLM to write code and another to review it is a circular trap, prompt engineering shouldn't be a computer scientist's day job, and every time you reject your code AI's output, you're quietly generating the training data that shapes its successor. Chapters 00:00 Introduction to Laura Dietz and Her Journey 03:12 Exploring LLMs as Judges 06:16 Challenges in Evaluating Search Systems 08:49 The Evolution of User Queries and Expectations 11:46 The Role of LLMs in Information Retrieval 14:44 Defining Quality in Search Results 17:27 The Complexity of User Intent 19:54 Human-AI Collaboration in Code Review 22:53 The Future of LLMs in Software Development 25:23 Balancing Human and AI Roles 28:20 Innovative Approaches to AI Evaluation 34:10 The…
People in this episode
Host: John Crickett
Guest: Laura Dietz
Topics covered
- AI evaluation
- LLM
- software development
- code review
- human-AI collaboration
- prompt engineering
Keywords
- AI evaluation metrics
- LLM as a judge
- software engineering
- training data
- code review
- human-AI collaboration
Mentioned in this episode
Organizations: computer science
More episodes of Coding Chats
- Startup Advisor Secrets: Hiring, CTOs & Going From POC to Product · May 28, 2026 · 24 min
- AI Agents Have a Memory Problem (And You're Probably Making It Worse) · May 21, 2026 · 47 min
- I got into computers to avoid people then they put me in charge of them! · May 14, 2026 · 49 min
- The Death of Writing Code: OpenAI's Engineer on the Rise of Harness Engineering · May 7, 2026 · 50 min
- Let it crash! How Erlang and BEAM build bullet proof software · April 23, 2026 · 37 min
- AI writes it. You own it. Don't ship AI slop · April 16, 2026 · 56 min
Explore listener stats, chart rankings, contacts and more on the Coding Chats podcast page.