Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
I am glad that Bernard is not engaging in alarmism about AI and recognizing the …
ytc_Ugwj5CTy4…
G
"Turns out it isn't bad for the environment either" tell that to the cities who …
ytr_Ugxp5AUlM…
G
34:50 Kinda a huge assumption that Elon Musk has no moral compass. Out of all th…
ytc_Ugwvx6Bgq…
G
Im imaging drunk people taking the food out and eating it then the using the ro…
ytc_UgzcdiIuG…
G
If you are not afraid; you did not understand it.
Only computer experts know wha…
ytc_UgzYf5000…
G
I'm glad they're suspending it besides they're going to get rid of Jobs with stu…
ytc_UgxlI4kAt…
G
But to whom will this ai companies sell when all other companies are making loss…
ytc_Ugxd7fWXx…
G
"pro-ai artists" are literally just lazy ass people not willing to put the time …
ytc_UgzbELDNT…
Comment
Bring Back Full Regeneration Access, Not Just GPT-4.0
So apparently GPT-4.0 is coming back. Great. But that’s only half the problem.
The regenerate feature is still crippled, and this needs to be reversed just as urgently.
Before, you could go back and regenerate any past AI reply, even if you were two or three messages beyond it. That mattered because not everyone catches a response they want to redo immediately. Sometimes you realize it later, and the ability to go back was a core workflow for a lot of us.
Now? You can only regenerate the very last AI message, or you have to dig through menus — and sometimes the option just disappears completely. This change is pointlessly restrictive. It doesn’t “streamline” anything, it just breaks established use patterns for no benefit whatsoever.
We already pay for Plus because we value the flexibility and control this app used to have. Stripping away that control, one “update” at a time, while charging the same price is a slap in the face to paying subscribers.
If OpenAI is listening:
• Bring back unrestricted regeneration for all past messages.
• Stop removing basic functions people rely on.
• Focus on fixing the problems we’ve actually been asking you to fix instead of rolling out silent downgrades.
If you agree, comment so this gets visibility — because GPT-4.0 coming back is great, but if regeneration stays locked down like this, they’re still taking away what we paid for.
reddit
AI Responsibility
1754694521.0
♥ 3
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | company |
| Reasoning | unclear |
| Policy | regulate |
| Emotion | outrage |
| Coded at | 2026-04-25T08:33:43.502452 |
Raw LLM Response
[
{"id":"rdc_n7kxusj","responsibility":"company","reasoning":"unclear","policy":"unclear","emotion":"resignation"},
{"id":"rdc_n7kyqmc","responsibility":"company","reasoning":"unclear","policy":"none","emotion":"outrage"},
{"id":"rdc_n7mbvuh","responsibility":"distributed","reasoning":"deontological","policy":"unclear","emotion":"fear"},
{"id":"rdc_n7oeqbd","responsibility":"company","reasoning":"unclear","policy":"unclear","emotion":"outrage"},
{"id":"rdc_n7orhhl","responsibility":"company","reasoning":"unclear","policy":"regulate","emotion":"outrage"}
]