A test is the act of exercising a system with test-cases, in order to find failures or demonstrate a correct outcome.
This process looks like:
- Test planning → Test case development → Running test cases → Evaluating test cases.
It is impossible to eliminate all Software Bugs and to test everything. Therefore, we need to wisely choose what we test do balance out the cost of testing.
Testing cannot verify requirements. Incorrect or incomplete requirements may lead to spurious tests. Bugs in test design or test drivers are equally hard to find.
How do you test GenAI software?