Natalia Petrovskaia: The AI Illusion: Rethinking the Role of Human Expertise in Software Testing
summary
Claims that AI will handle routine testing but not replace humans, citing reports that AI usage in testing will rise almost 40% by 2030 and that more than half of companies already use ML/AI. Demonstrates tools and failures: AI generates test cases and data quickly but also hallucinates (invented libraries, fake papers) and ChatGPT 3.5 hallucinated; an experiment produced 70 checks vs 110 test cases from humans in 1.5 hours. Recommends combining AI for speed with human system thinking, intuition, creativity and empathy.
AI-written summary · about
chapters (10)
- 0:00Human expertise in AI era
- 1:11AI’s role in testing
- 3:18AI tools strengths and limits
- 8:17Tasks AI cannot handle in testing
- 10:50System thinking for testing
- 15:08Intuition in testing
- 21:12Creativity beyond AI
- 27:59Empathy-driven testing
- 32:02Human expertise with AI
- 35:39Q&A
related talks
44mHow AI is Exposing the Existential Crisis of Software Testing
46mThe Real Impact of AI in Software Testing
01h 01mRise of the Machines: Can Artificial Intelligence Terminate Manual Testing?
40mShould We Let AI Take Over Test Automation Completely?
32m🧐 A Look at Possibilities of Using AI in #SoftwareTesting
23mHuman-Driven Intelligence: Where AI & Automation Fall Short
47mHuman Experience Testing – Redefining Quality Engineering in the Age of AI
44mAI Empowered Software Testing