I’m a tester. That means I spend a good chunk of my time politely informing developers that their “minor tweak” has just set fire to half the application. Glamorous, I know. So, naturally, people ask me: “Aren’t you worried AI will take your job?”
To which I say: “Only if AI enjoys interminable meetings, inconsistent requirements, and a dev who insists it works on his machine.”
But seriously, no, I’m not worried. And I’ll tell you why. Because even the smartest AI still needs a babysitter. A tester. A sentient being capable of nuance, context, and spotting that the “Cancel” button has accidentally been labelled “Launch Missile”.
AI isn’t a threat. Brilliant at some things, hopeless at others, and constantly in need of supervision.
Just look at all the disclaimers. “May hallucinate.” (this is a genuine phenomenon!). That’s not a feature; that’s a warning you usually reserve for dodgy mushrooms or Eurovision entries. If we shipped software with the same success rate as some AI models, I’d be testing toaster firmware by now.
So, what do I really think about AI? I think it’s brilliant. Genuinely. It’s made me a better tester. It’s like having an assistant who never sleeps, is never 'away' on Slack, and can explain a regex without crying.
12 ways AI Has Made Me Look Smarter Than I Am
1. Software Wizardry for the Uninitiated
At its most basic, I use AI to help me understand unfamiliar tools or concepts. The first time I checked security protocols in DevTools, I accidentally opened a wormhole to 2006. AI gave me the step-by-step like a patient satnav: “Turn left at Network tab, continue to Security, and try not to cry.” Step by step instructions for most commercial software can be had and this helps me in two ways - triggering ideas of ways to test those pathways but also those step-by-step instructions can become "test-step" by "test-step" test cases.
2. ^[^\s@]+@[^\s@]+\.[^\s@]+$? Regex? Sorted.
Regex - either you love it, or you find yourself deep in a Stack Overflow rabbit hole, questioning your life choices.
AI’s brilliant at breaking down complex regex patterns into something that actually makes sense. I've used it to generate valid and invalid strings, so I can quickly add them to the manual and automated versions of the scripts I am writing (that's an email regex by the way!)
3. Edge Cases? Filed.
Need to uncover obscure edge cases? One of the most obvious ways a tester operates is to push the boundaries. AI helps me identify boundary values and the sneaky outliers that lurk in the shadows, waiting to break things so that I can make sure to script scenarios to catch them. This is particularly helpful for larger datasets where AI can digest and interpret it much faster than I could.
4. Mock Data on Tap
Following on from edge cases, when I need dummy data AI is as fast as any online generator but saves me from having multiple browser plug ins to slow down my already overworked browser. Names, NI numbers, postcodes, email addresses that look just real enough to cause concern? AI’s got more identities than I've got browsers installed. And all of them are clean, structured, and don't trigger any GDPR nightmares.
5. Test Flows? Mapped.
When a process diagram looks like it was drawn by a toddler during turbulence, AI helps straighten it out. It can translate those spaghetti-like flows into clearly testable steps and turn chaos into clarity. Identifying all of the critical pathways is a key component of test creation, a task that AI handles accurately and swiftly.
6. Accessibility Spotter with Laser Vision
AI can’t run my browser tools (yet, that I am aware?), but it can tell you what to look out for (and as in point 1, where to look for it): contrast issues, focus traps, missing ARIAs. Basically, it helps me write tests that help make sure that our sites actually work for everyone, not just those with superhuman eyesight and mouse accuracy.
7. Exploratory Testing Compass
I tell it what I'm testing, and AI will highlight areas I might've overlooked - dodgy state transitions, hidden inputs, gremlins that only show up when Mercury’s in retrograde. I poke it, it pokes back with insight. All the while improving my testing and educating me along the way.
8. Bug Ticket Peace-keeping Force
I genuinely respect our Mando Group developers, they are all truly helpful, lovely people, so AI has helped me keep things civil and helped me to maintain a good working relationship! Let me explain: at times, my sometimes irritational human emotions come to the fore, usually when deep in the testing weeds and getting a rejected bug with what I perceive as a lazy response: "is this still happening?"- so, when I need to write the bug report that’s clear but doesn’t scream “I’m passive-aggressively fuming”, AI’s got my back. It helps me phrase it just right - sharp enough to be taken seriously, smooth enough to avoid war.
9. Translate Dev-Speak into English
Similar to point 8, a Dev says: “It’s just a race condition between asynchronous DOM updates and lazy-loaded resources.”
AI says: “The page’s loading like it’s drunk. Try this.”
Finally, someone who can translate developer wizardry into something the rest of us mere mortals can act on.
10. Explaining Code I Didn't Write (or Understand)
Ever been handed code by a developer who’s since “moved on”? AI’s like having a translator who doesn’t judge you for asking, “What in the name of C# is this doing?”. It is a such a time saver to be able to have complex methods explained quickly and plainly and this can also be used by me to then refactor and simplify that complex code when needed.
11. Boilerplate Code Generator Extraordinaire
Creating the mundane, the setup/teardown, data providers, page object scaffolding… the tedious automation necessities! AI doesn’t complain. It’s like having a clone who loves writing the boring bits and I am on hand to review, guide and verify the outcome leaving me to write the more interesting parts of the code.
12. Selenium Debugging Pal
AI will sniff out all the usual offenders: flaky selectors, redundant waits, basically, it’s like a code reviewer with no filter and infinite patience polishing my code. AI can help me brainstorm alternatives - whether it’s using XPath like a scalpel, hopping into iframes like a test ninja, or breaking out the JavaScript hammer. It's like having a troubleshooting mate who I don't have to feel guilty about bugging 5 times an hour!
So there we have it, AI hasn’t replaced me, it’s enhanced me. It’s made me quicker, sharper, and possibly more tolerable in meetings. It’s a tool, not a threat. A hyper-intelligent spellchecker with a flair for regex. It takes the dull, the complex, and the just-plain-annoying, and turns it into something manageable. It doesn’t replace what I do - it amplifies it.
And if one day AI does come for my job?
Well, I’ll leave it some delightfully vague acceptance criteria and see how it gets on - but if it ever learns the "testers sigh" or infamous eye-roll, THEN I might be out of a job!
Want to receive more expert insights from our team?
