GenAI for Realistic Test Data, Edge Cases, and Better Creativity
summary
Demo shows GPT-3.5 prompt→code+unit tests producing a factorial function and additional tests for positive, negative, string and non‑integer inputs, then finds a generated edge case that triggers recursion/overflow and rewrites the code iteratively to avoid stack overflow. Also generates tabular positive/negative/edge test scenarios for an Amazon purchase flow with high/medium/low automation priority and uses Bard, DALL·E/Midjourney, GitHub Copilot, Hugging Face and Google Project IDX to create images, prototypes and continuous‑testing helpers while noting bias, privacy and resource limits.
AI-written summary · about
chapters (3)
related talks
48mTestGPT: Generative AI (AutoGPT, ChatGPT & VisualGPT)
43mBecome a 10x developer by harnessing the power of Generative AI
41mAI-Augmented Testing: How Generative AI, Prompt Eng Turn Testers Into Superheroes
51mOpenAI and the AI that may overtake humans
36mUnleashing Generative AI: Improving Developer Productivity in SRE
40mDavid Fogl: Leveraging Generative AI for Enhanced Software Testing
16mCan Test Driven Development be speeded up with Generative AI?
38mIntroducing the New Colleague - ChatGPT