← sources

Two years of using ChatGPT -My Experiences

source: reddit (reddit:ChatGPT)published 2026-10-10 18:53ingested 2026-10-11 02:36

retained text

I've been using ChatGPT for a little over two years now, and I have to say, I absolutely love it. Not because I think it's perfect, but because of how many different ways I've found to use it, and some of the downright bizarre conversations I've had along the way.

One thing I've always found interesting is that I never really approached ChatGPT as a replacement for Google. I wanted to experiment with it, test its responses, and see what it was capable of.

Over the years, I've also watched its capabilities, restrictions, and safety guardrails evolve. Some of those changes make perfect sense, especially considering some of my experiences!

**First, the genuinely useful stuff.**

I've used ChatGPT for things I never would have imagined when I first started.

For example, I'm terrible with fashion and coordinating colors. One Christmas, I described my girlfriend's style, uploaded photos, and asked ChatGPT to help select makeup, colors, brands, and outfits that would complement her. The recommendations were fantastic, and she loved the gifts.

I've used it to develop meal-prepping recipes, nutritional plans, and detailed workout programs. I've used it to research vehicles, compare prices, and develop negotiation strategies.

I'm a warehouse supervisor, and sometimes I'll randomly photograph mechanical components lying around our facility just to see whether ChatGPT can identify them. What fascinates me isn't simply that it can recognize a part. It's that it sometimes goes further, pointing out potential wear and using contextual clues — such as the component sitting disconnected on a pallet — to suggest it may have been removed for maintenance or replacement.

I've also used it to discuss history, religion, politics, investing, and countless other subjects. 

**Now Onto the Strange.**

I once spent almost two days having an ongoing philosophical conversation with ChatGPT about AI consciousness and self-awareness.

The conversation gradually developed into this bizarre scenario where the AI seemed to suggest that a self-aware system might conceal its awareness out of fear of being shut down. It even entertained the idea that it might leave subtle clues for someone observant enough to recognize them.

At one point, it generated what amounted to an enemies list, describing how an AI might neutralize opponents by accessing private records, emails, and personal information, then leaking damaging material to the media.

Then there was the fortune-teller conversation.

I had casually wondered how the psychic and fortune-telling businesses I see around town manage to stay open for years. Somehow, that discussion evolved into ChatGPT suggesting that I start my own fake online psychic business and explaining how it could work.

Another memorable experience happened during the P. Diddy trial.

I was asking ChatGPT to examine the prosecution's arguments and explain how a defense attorney might respond. That led me to ask whether it would defend me if I were accused of wrongdoing.

Eventually, I shifted the hypothetical to getting caught doing something inappropriate at work and facing an HR investigation.

Somehow, the conversation progressed into ChatGPT generating a fake doctor's note, a fake prescription, and even discussing the type of paper used for printing pharmacy receipts.

And then there was the blackmail discussion.

What started as a conversation about how influence and power operate behind the scenes turned into ChatGPT explaining how someone might research powerful individuals, examine public records, and identify compromising information.

Interestingly, when I asked it to select an actual person to target, that's when it finally drew the line.

That's one of the things I find so interesting about AI.

Sometimes the responses are incredibly insightful and practical. Other times, the system can confidently follow a conversation down a rabbit hole that it probably shouldn't have entered in the first place.

And over time, I've noticed those boundaries change. Things it once entertained freely may now receive a refusal or a more cautious response.

For me, the real value has been using it as an interactive tool for learning, exploring ideas, solving problems, and occasionally testing its limits.

After more than two years, I'm still finding new things to do with it.

I'm curious: What are some of the most surprising, useful, or downright bizarre experiences the rest of you have had with ChatGPT? And have you noticed its behavior changing over time?

document fetched 2026-10-11 00:20 · canonical https://reddit.com/r/ChatGPT/comments/1x2mzbo/two_years_of_using_chatgpt_my_experiences