I've been in this sub for 2 years and i want to name what it's become.
Scroll the front page any given day. Three categories dominate:
- "Look at this funny thing Cha͏tGPT said" (90% of which is the user leading the model into a gotcha)
- "ChatGPT got this obvious question wrong" (usually a question it didn't get wrong, but the OP misread the answer)
- "Is anyone else noticing ChatGPT got worse" (weekly, every week, for 18 months)
Almost zero content about people actually using AI to do real work. Almost zero concrete use cases. Almost zero "here's the 6-month log of an agent running my business." It's all reaction content.
I'm not on this sub to defend ChatGPT. I canceled Pl͏us in February. I'm here to point out that this sub has become a mirror for exactly the kind of person who was never going to make AI useful to their life. If the interface is a chat tab, and the model forgets you between sessions, and you have to type every request, and you can't connect it to your tools, then yes, you will eventually run out of interesting things to do with it and fall back to screenshotting its failures.
Meanwhile, look at the subs where people are actually running age͏nts. Concrete dated logs. Specific numbers. "My agent caught a missed invoice and paid for 3 months of my platform in one afternoon." "Pulled Str͏ipe every morning for 60 days, here's what it flagged." This content exists. It's not on this sub. It's on the agent subs and in places like RunLobster (OpenClaw) discussions.
My polarizing claim for this sub: the "ChatGPT is getting worse" narrative is mostly cope from people who were never actually getting real work out of it. The people who are getting real work out of AI don't use ChatGPT, they use agents. And they're not complaining here because they're too busy shipping.
If i'm wrong, prove it with your own 60-day ChatGPT log showing specific work done. I'll read every reply.
[link] [comments]