- Primary Subject: Fallout 4
- Key Update: Former tester says AI cannot replicate the chaotic creativity human QA testers use to break games
- Status: Confirmed
- Last Verified: March 16, 2026
- Quick Answer: Former Fallout 4 QA tester Colin McInerney says AI cannot replace human testers because real bug discovery often requires “professionally stupid” experimentation.
Former Fallout 4 QA tester Colin McInerney has explained why he believes artificial intelligence still cannot match the value of human testers, especially when it comes to uncovering the kind of strange, unpredictable problems that often appear in massive open-world games.
Speaking about his time working at Bethesda during the development of Fallout 4, McInerney said quality assurance was never just about following a checklist or replaying the game the “correct” way.
For him, some of the most valuable testing came from pushing the game into absurd situations, experimenting far beyond normal gameplay, and asking what would happen if he deliberately tried to break it.
How Did the Discussion About AI and QA Testing Start?
While AI can be useful for structured tasks or repeated testing routines, he argued that it lacks the instinctive curiosity and reckless creativity that often lead to meaningful bug discoveries.
In his view, testing a game properly does not always mean playing as intended. Sometimes it means ignoring the expected path completely and throwing the game into situations no designer planned for, simply to see where it falls apart.
That mindset, he suggested, is exactly why human testers remain so important. To illustrate his point, McInerney recalled one particularly chaotic Fallout 4 testing session on Xbox One, where he became interested in how close he could push the game toward its memory limits.
Since the console had 8GB of RAM, he started monitoring memory use and treating the test almost like a challenge.
Instead of following standard QA procedures, he tested the game by deliberately pushing it to its breaking point.
That experiment eventually pushed him to use console commands to boost his character to extreme levels and arm himself with weapons designed to cause maximum chaos.
What Does “Professionally Stupid” Mean in QA Testing?
Using that example, McInerney pointed to the difference between machine logic and human behavior, explaining that AI may be useful in development but usually stays within predictable logic, unlike human testers who experiment in strange or ridiculous ways.
That is where his now widely shared comment came from, when he joked that he had to be “professionally stupid” in ways a machine never could.
What he meant was that good QA sometimes depends on doing the dumbest possible thing on purpose, because players eventually will do it too.
Human testers can imagine bizarre scenarios, combine systems in unintended ways, and follow chaotic instincts that do not necessarily come from logic, but from experience, curiosity, and a willingness to be reckless with the tools they are given.
That idea also struck a chord with players and developers alike, as it reflects a long-standing truth about games: people rarely play them the way designers anticipate.
If a mechanic can be abused, stretched, stacked, or turned into something ridiculous, someone will eventually do it.
That is why QA is so often about more than checking whether quests trigger correctly or menus display the right text.
It is also about finding the weird edge cases, the extreme combinations, and the unexpected interactions that can cause a game to glitch, crash, or break entirely.
McInerney’s story is a dramatic example of that process, but it also captures the larger purpose of QA in general.
The job is not only to verify that the game works when played normally, but to anticipate the kinds of nonsense real players might attempt once the game is out in the world.
For more like this, stick with us here at Gfinityesports.com, the best website for gaming news.