Ask HN: How are teams validating AI-generated tests today?
Posted by sriramgonella 8 hours ago
With the rise of AI-assisted development, many tools generate tests automatically.
But validating whether those tests actually cover meaningful edge cases seems harder.
Curious how teams here handle this in real workflows.
Comments
Comment by david_iqlabs 8 hours ago
I've found it works better when the AI is just explaining results that come from deterministic metrics rather than inventing the analysis itself.
Curious how other teams are dealing with that.
Comment by sriramgonella 8 hours ago
Comment by david_iqlabs 42 minutes ago
I spent months trying to make an executive narrative generated by AI, but eventually moved away from that approach. The results were often inconsistent or overly generic, which made it difficult to rely on the output for serious reporting.
In the end I shifted to a fully model-driven approach where the narrative is built directly from structured signals and scoring logic. That made the reports far more accurate and evidence-based, and it keeps the output consistent from scan to scan.
Comment by itigges22 8 hours ago
Another counter-measure I have is to simply lock code before testing. Look over test files, and ensure its not following the happy path.
Comment by sriramgonella 7 hours ago
Comment by itigges22 2 hours ago