---
title: "Comment by Michael Weber on \"Agentic AI Can Write Your Test Cases. But Can It Think Like a QA?\" | Hashnode"
description: "Fully agree. As Michael Bodnarchuk recently pointed out, the real challenge isn't just generating code, but ensuring its maintainability. I’ve been looking into"
url: "https://hashnode.com/posts/agentic-ai-can-write-your-test-cases-but-can-it-think-like-a-qa/69d87033fa7251682e3ebd02/comment/69d893edfa7251682e5e4f3d"
---

[Discussion](https://hashnode.com/posts/agentic-ai-can-write-your-test-cases-but-can-it-think-like-a-qa/69d87033fa7251682e3ebd02)

Comment [Agentic AI Can Write Your Test Cases. But Can It Think Like a QA?](https://hashnode.com/posts/agentic-ai-can-write-your-test-cases-but-can-it-think-like-a-qa/69d87033fa7251682e3ebd02)

[Michael Weber](https://hashnode.com/@weberstmichael)

[Apr 10](https://hashnode.com/posts/agentic-ai-can-write-your-test-cases-but-can-it-think-like-a-qa/69d87033fa7251682e3ebd02/comment/69d893edfa7251682e5e4f3d)

Fully agree. As Michael Bodnarchuk recently pointed out, the real challenge isn't just generating code, but ensuring its maintainability. I’ve been looking into how ai in software testing can actually solve this by focusing on orchestration rather than just raw script generation. It’s a game-changer for long-term stability: [https://testomat.io/blog/ai-in-software-testing/](https://testomat.io/blog/ai-in-software-testing/)

[Vishva Maithri](https://hashnode.com/@vishvamaithri)

[Apr 12](https://hashnode.com/posts/agentic-ai-can-write-your-test-cases-but-can-it-think-like-a-qa/69d87033fa7251682e3ebd02/comment/69d893edfa7251682e5e4f3d#reply-69db6405aadf1107e2b304f2)

Good point on maintainability! That's definitely a challenge teams face as AI-generated test suites grow. Orchestration over raw generation is a solid direction.

That said, I think the bigger shift is in how we think about the role itself. Even with great orchestration, someone still needs to bring the domain context, the edge-case instinct, and the quality culture to the team. That's the QX Engineer layer that sits above any tooling.

Appreciate you reading and sharing your thoughts!

[Michael Weber](https://hashnode.com/@weberstmichael)

[Apr 16](https://hashnode.com/posts/agentic-ai-can-write-your-test-cases-but-can-it-think-like-a-qa/69d87033fa7251682e3ebd02/comment/69d893edfa7251682e5e4f3d#reply-69e084de05f9fd0063636b89)

Absolutely, Vishva! I love that "QX Engineer" term—it perfectly captures the shift from 'writing scripts' to 'designing quality.'

You're 100% right: no matter how good the orchestration is, the "edge-case instinct" is purely human. Tooling can handle the repetitive execution, but it's that domain context and culture that prevent the system from just efficiently running the wrong tests. It’s all about empowering the engineer to focus on the high-level strategy rather than getting bogged down in the mechanics.

Appreciate you adding that layer to the discussion!
