.NET Rocks! AI Concerns with Mark Seemann
Jul 24, 2025
Mark Seemann, a programmer with an economics background and blogger at blog.ploeh.dk, shares his insights on the impacts of AI, particularly large language models, on software development. He addresses the pitfalls of viewing AI as an ultimate source of truth and the essential knowledge required to evaluate AI-generated code. The conversation touches on the challenges faced by junior developers, the ethical implications of AI, and the importance of critical thinking in an era increasingly reliant on technology. Practical discussions around integrating AI tools highlight the need for careful oversight.
AI Snips
Chapters
Books
Transcript
Episode notes
Blind Trust in AI Answers
- Many people treat AI generated responses as ultimate truth without verification.
- This blind trust in AI outputs is a new risky behavior impacting software development.
ChatGPT Product Recommendation Fail
- Carl Franklin tested ChatGPT for product recommendations but found it made inaccurate suggestions.
- He verified the product reviews himself and returned a product that didn't meet his needs.
Verify AI-Generated Code Carefully
- Never rely on AI-generated code without verifying and understanding it yourself.
- If you lack the knowledge of the target language, reviewing AI code is risky and likely to fail.
