Welcome to the 323rd issue!

Looking for ways to upskill in AI for testing?

Today, I want to share with you two great collections:

• Awesome AI Testing by Tuğkan Boz
• AI Learning Plan for Testers by Rahul Parwal

Both are full of tools, free courses, blog posts and other resources.

Happy testing! 🙂

Dawid Dylowicz

Does AI change the value of the expertise we've developed over the years? Drexel McMillan gives a thought-provoking view on the impact of AI on our roles and worth.

On that note, Emna Ayadi shares a reflection in My Emotional Shift with AI, while Aryadevi Neelakantabhattathiri explains What vibe coding taught me about the future of QA.

Wondering how to strengthen quality culture on a team with no dedicated QA? Maria Kourtesi shares a year-long journey of building trust and progressing with small wins of influence.

Jitesh Gosai shares lessons learned from how even a solid regression test setup can fall apart over time once the people who built it move on.

Similarly, Fred Hebert looks at the bigger picture in Control and complexity: tension in systems design.

Creating software with AI is easy. Making sure it works? Much harder. Jason Arbon emphasises why we should get better at validating what it builds, followed by another piece of advice — Don't Just Use AI to Build Faster. Use It to Ship Faster.

Some companies may decide to reduce the size of their QA teams. Alden Mallare breaks down what this means from a quality-ownership perspective and how we can stay valuable as testers.

Moreover, here's an interesting discussion on Reddit about the situation where Management is nudging QAs to become hybrid QA/Dev.

Johann Eicher explains step-by-step how they leveraged AI to analyse and greatly reduce the test suite's execution time without changing any test logic or assertions.

At the same time, Rajeshkumar Rajaseakaran Nair tells us Why AI Testing Needs Confidence Scores, Not Just Pass/Fail Results.

James Thomas splits test data into three kinds: for specific logic, for performance checks and for finding new states in an app, each requiring a different approach.

Similarly, Irfan Mujagić explains Why Your Golden Dataset Has a Shelf Life.

How can you trust an AI-generated test? Vladyslav Dmitriiev built a pipeline with a reviewer, a judge and mutation tests where every step is checked by either a different model, a script or a person.

On that note, Pranav Pandit advises on When AI maintains your tests: Designing safe self-healing automation.

Wondering how Amazon solves testing a system of 1,000s of interconnected services? Carlos Arguelles explains five pillars for running tests in production.

Also, Rajeshkumar Rajaseakaran Nair points out The Biggest Mistake Test Engineers Make After a Regression Test Fails.

Modern Android testing solutions, like Compose and JVM-based rendering, challenge the traditional testing pyramid. Mahmoud Ramadan explains why the testing trophy model fits better for testing the behaviour.

Similarly, Mayvin Ramasawmy starts a new series on mobile testing: Parallel Mobile Testing, Part 1: Make Every Scenario Truly Independent.

Want to speed up your Playwright test runs? Andrey Lushnikov explains how WordPress cut its shard time from 35 to 23 minutes using a new balancing reporter.

Also, Gleb Bahmutov explains How to stop or skip Cypress tests without losing the information.

Bruno — a popular API test client — has recently added AI features, such as autocomplete, test and doc generation, that you can power with your own LLM.

Also, if you're doing API performance testing, NaveenKumar Namachivayam has Built an MCP Server for JMeter Docs.

Playwright ships with eight different reporters and Svetlana Tretjakova breaks down when each one is useful, from simple terminal output to JSON, JUnit and blob reports for CI pipelines.

Additionally, Katsuya Oura explains Graduating from toBeTruthy: Choosing Meaningful Matchers in Playwright's expect.

Birgitta Böckeler tried three spec-driven development tools: Kiro, spec-kit and Tessl, to see how writing specs before implementation works in practice, along with the problems this approach runs into.

Moreover, Uberto Barbini takes a more critical view: Spec-Driven Development: The New Waterfall.

A good, high-level refresher by Daniel Knott on what DevOps is and what solutions testers should get familiar with to strengthen the shift-right testing strategy.

In this insightful 40-minute discussion, Andrew Knight shares how their team uses AI coding agents, spec-driven development and Playwright to keep up with increasingly faster development cycles. Hosted by Joe Colantonio.

Code reviews... 😂

Thanks for reading!

If you enjoy this newsletter and find it helpful in becoming a better tester, please consider sharing it with others.

Dawid Dylowicz