Unit Testing with AI – Webinar Recording

Welcome back to my “…with AI” series! In this latest session, we’re diving deep into a topic every developer grapples with: unit testing. But we’re giving it a 21st-century spin by bringing AI into the mix.

We’ll explore how AI-powered unit testing is not just about isolated functions anymore.It’s also about component testing. I also explain why component tests fit into the unit testing drawer.

The Power of Context in AI-Assisted Unit Testing

The single biggest challenge in getting an AI to do exactly what you want is context. When you ask it to generate code or tests from a distance, it’s often guessing.

But with unit testing and component testing, the rules change. The AI is close to the code; it has direct context. When we combine this with another layer of context we can provide – in my example, a known algorithm – AI can do more than just help. It can provide massive leaps forward in our testing workflow, generating test cases in seconds.

Some even run.

You are the Gatekeeper: The Human in the Loop

So, AI generates a suite of perfect unit tests, and our job is done, right? Not so fast.

All we’ve got left is to approve it, and that’s not a small feat. It’s a huge responsibility. We, the developers, are the ultimate gatekeepers. We must decide if the AI-generated tests are good enough, comprehensive enough, or if they even run. (Spoiler: Sometimes they don’t.)

This session shows that while AI can give us a powerful head start, the human in the loop must be the validator, the approval master. Without this critical human oversight, we’re left with tests we either don’t trust or don’t understand. And code that can’t be trusted is a liability.

We are the final checkpoint for quality.

What’s Your Take?

So, over to you. Do you use AI to generate tests for your code? What do you think about the quality of these AI-powered tests? Do they work for you, and how do you ensure they’re up to snuff?

Share your thoughts in the comments below!