There is no I in Al
summary
Reviews AI approaches for testing and introduces an open‑source tool called Alumnium that exposes an AI-powered API for Selenium as a small Python library. It runs against Entropic, Google Gemini, OpenAI and open‑weight models like Llama/deep seek, reports cost around $3 per 1,000 tests, and shows demos (calculator, demo shop) that pass. Performance measures show 3s → 50s (cache → 3s); current limits: Chromium accessibility only, mobile and Firefox/Safari support pending.
AI-written summary · about
chapters (11)
- 0:00Introduction
- 1:24Speaker background
- 2:07Page objects approach
- 6:44Frustration with complexity
- 10:15Introducing Aluminium
- 15:10Demo with todo app
- 22:18Cost privacy AI options
- 22:38Live demo calculator tests
- 26:32Limitations and future plans
- 29:47Call to action
- 30:08Q&A
related talks
55mAI-First Testing
43mAI for Testing – Panel Discussion
41mTest Automation in the Age of Al
41mAI Driven Mobile Testing
15mAlgorithms That Run AI and How to Use It in Testing
01h 16mAI in Software Development: Boosting Efficiency
32mQuality engineering in the AI age: From testing AI to leveraging GenAI
44mAI Empowered Software Testing