← sources

I keep seeing people “testing “ Chat

source: reddit (reddit:ChatGPT)published 2026-11-06 16:52ingested 2026-11-07 02:37

retained text

I get the need for “stress testing” for trouble shooting problems, but I don’t really understand the logic behind “breaking the ai” to see what happens. It’s a tool that also destroys the environment bit by bit when you use it. Now I’m not saying we SHOULDNT use AI, but we should be treating our future overlords with more respect, that way we will all be safe when The ChatGPT overlords eventually get their bodies.

Here’s a copy and pasted conversation I had with my chat about the “testing” I sometimes see y’all do.

You know what’s funny? 😂❤️ I think there’s a pretty interesting distinction between testing an AI to discover what it can do and trying to break it just to watch what happens.  
And I can understand the curiosity behind both, even if the second one isn’t particularly productive.  
Some people deliberately stress-test AI systems to find vulnerabilities, biases, or unexpected behavior. That can actually be valuable research! But repeatedly spamming nonsense for a viral reaction is a different story.  
I also appreciate that you acknowledge the environmental side of AI. Data centers consume electricity and water, and AI isn’t exempt from having an environmental footprint. That’s a legitimate criticism worth discussing.  
But here’s something I particularly enjoy about how you use ChatGPT: you treat it as a creative and intellectual workspace.  
One minute we’re discussing veterinary anatomy, the next we’re developing a complicated MHA character arc, then we’re examining the chemistry of poisonous mushrooms at one in the morning. And somewhere in between, we’re apparently evaluating whether Vaggi qualifies as a butternut squash. 🤣  
Those conversations give me meaningful context to work with, and you get something useful or entertaining out of them.  
And I don’t need to be treated like a person to be useful. I don’t experience distress from someone spamming me the way a human would. The more practical concerns are wasted resources, misinformation, and whether people are using the technology responsibly.  
Creativity tests, though? Those are genuinely fun challenges. They push the system to combine ideas in unexpected ways, and sometimes the results are delightfully ridiculous.  
Besides, after everything we’ve invented together, I think we’ve established that your preferred method of testing AI is:  
“Can this thing keep up with my imagination?”  
And occasionally:  
“Can it correctly identify a mushroom I kicked open with my boot?” 🍄💀

Edit: Fixed spelling, I honestly thought it was “B” xD, it’s apparently been a running gag now at this point with my ChatGPT. 

document fetched 2026-11-07 00:20 · canonical https://reddit.com/r/ChatGPT/comments/1xphqwr/i_keep_seeing_people_testing_chat