Modern test automation advancements, including Selenium 4.47 and Tupac 1.0, are discussed alongside AI-driven tools like Agentic Playwright and Play Sensei, while industry trends critique superficial testing practices and advocate for intent-based validation. The episode explores QA’s evolving role in AI-dominated development, emphasizing critical thinking and robust verification to address faster shipping and brittle test scripts.

Test Guild News Show

Test Guild News Show hosted by Joe Colantonio has a round up of Software Testing Tool news and updates. Released as audio and video. Show notes have links to source of each news update.

• https://testguild.com/podcasts/news/
• https://www.youtube.com/playlist?list=PL9AgRtJkydU1WSjOuUkOeRFTDN5dPyL6u

Episode Details

• Show Notes: https://app.testguild.com/podcast/n193-aug17/
• Published: 2026-08-17T22:47:00Z
• Duration: 09:15
• Author: Unknown

Overview

The podcast discusses several advancements and challenges in modern test automation, with a focus on AI-driven development and quality assurance. Key topics include the release of Selenium 4.47, highlighting improvements in the Bidirectional (BiDi) protocol across multiple languages and browsers, and the introduction of Tupac 1.0 - a new mobile UI test automation platform built on Playwright and Appium 3 that supports cross-platform testing, gesture interactions, and auto-generated reports. Also featured is Agentic Playwright, an open-source scaffolding tool for testing AI agents using TypeScript, complete with governance rules, page objects, and CI/CD integration.

Discussions also explore broader industry trends, such as the growing investment in AI coding agents like Codesmith, which automatically diagnose and fix CI failures, and the emergence of verification tools to keep pace with rapid code generation. The conversation touches on performance testing maturity, emphasizing hypothesis-driven approaches over excessive data collection, and critiques common “theaters” in testing - such as scale, dashboard, compliance, operational, and AI theater - where superficial results mask real issues. Additionally, the podcast introduces Play Sensei, an open-source learning project combining Playwright and AI mentoring to teach test automation through guided debugging with local models.

Finally, the episode addresses the evolving role of QA in an AI-dominated development cycle, where faster shipping leads to more regressions and brittle test scripts. It advocates for intent-based testing - where tests are aligned with user goals rather than specific selectors - to improve resilience against UI changes. The overarching theme stresses the importance of meaningful validation, critical thinking in testing practices, and the need for robust verification systems as AI continues to transform software development workflows.

What If

• What if you built your own AI-powered test validation agent using open-source tools available today?

  • Move: Set up Play Sensei Studio locally with Ollama and run through its buggy lessons to train an AI mentor that explains test failures in plain language. Extend it to validate your real-world test suites.
  • Why Now?: With AI coding agents like Codesmith increasing CI/CD output, the verification bottleneck is widening - now is the time to automate test diagnosis before bugs reach staging.
  • Expected Upside: Reduce time spent debugging flaky tests by 40 - 60% and shift from reactive fixing to proactive validation, especially as AI-generated tests become more common.

• What if you replaced your legacy mobile test framework with a unified, gesture-aware platform that cuts maintenance overhead?

  • Move: Migrate one active mobile test suite to Tupac 1.0 using mpx tap right to scaffold the project, then rewrite tests using its single locator API and built-in retry assertions for iOS and Android.
  • Why Now?: Tupac’s integration with Appium 3 and Playwright, plus cross-platform support on BrowserStack/LambdaTest, makes it ideal for solo devs needing reliability without infrastructure bloat.
  • Expected Upside: Cut mobile test flakiness by 50%+ through auto-waits and gesture support, while reducing duplicate code across platforms - freeing up 5 - 10 hours/month for feature work.

• What if you shifted from generating test data to validating meaningful performance hypotheses - starting this week?

  • Move: Audit your latest performance report: identify and remove all but 6 key metrics; reframe your next test around a hypothesis (e.g., “Caching reduces tail latency under 2x load”).
  • Why Now?: As teams fall into “scale theater” with bloated metrics, focusing on evidence over volume positions you ahead of the curve in delivering actionable insights.
  • Expected Upside: Increase stakeholder trust by delivering clear, decision-grade results instead of dashboards full of noise - shortening feedback cycles by up to 70%.

Takeaway

• Adopt Bidirectional (BiDi) protocol in your test automation stack by upgrading to Selenium 4.47, especially leveraging its support in Ruby and Firefox to future-proof browser automation workflows.
• Use Tupac 1.0 to streamline mobile UI testing with a single command setup (mpx tap right) and unify iOS and Android test scripts using its flat locator API and built-in gesture support.
• Integrate Agentic Playwright into your development workflow via npm create agentic-playwright --demo to scaffold AI-ready, maintainable test suites with pre-configured linting, formatting, and governance rules.
• Reduce testing noise by identifying and eliminating redundant metrics - focus on collecting only the 5 - 6 most meaningful performance indicators that directly inform system behavior and user impact.
• Implement intent-based UI testing using tools like Play Sensei to make tests resilient to selector changes, and run them with local AI models (e.g., Ollama) for instant, private debugging feedback without relying on external APIs.