From Entertaining to Useful: LLMs for Test Tooling
summary
Claims that LLMs move from entertaining to useful with Cursor and Claude 3.5 Sonnet are backed by concrete builds: a QR code duplicate detector browser plugin, a session-extender, a JMeter listener, a Notepad++ personal-number scanner and a SOAP test-data generator. Delivery is fast—about one small tool per week—often usable on first or second try, sometimes buggy or hallucinating, and requiring exploratory testing and unit checks before trusting results.
AI-written summary · about
chapters (12)
- 0:00Specialized testing tools
- 3:45LLM evolution
- 6:41Custom testing tool
- 12:02Tester tool categories
- 25:07LLM coding lessons
- 31:50LLM syndromes
- 34:20Possibilities and drawbacks
- 36:08Summary and future
- 37:17Q&A reporting
- 39:01Q&A tool selection
- 40:06Q&A sensitive data
- 40:50Q&A usability
related talks
44mPractical LLM Evals: Down-to-Earth Tooling for Everyone
35mEvaluating LLM systems: the nuts and bolts
38mMaking LLMs reliable - A practical framework
28mGetting More Out of LLMs for Ruby Developers Justin Wienckowski, Jake Gribschaw We're all hearing about LLM tools that are changing on a daily, if not hourly, basis
01h 15m10x Development: LLMs For the working Programmer
44mAccelerating Software Development: The LLM (R)evolution
15mLLMs and Test Automation: A New Era of Self Healing
33mExplore LLMs and Multimodal Approaches in Test Automation