Raw LLM Responses
Inspect the exact model output for any coded comment.
Look up by comment ID
Random samples — click to inspect
G
Ai is not great at exaggerating movement and expression, use this tip to find ai…
ytc_UgznJtLje…
G
Yes, but.. question is, what people gonna do?
Almost every professions can becre…
ytc_Ugztbt822…
G
Sounds like "how can we get a cut of our share of profit from this whole AI thin…
ytc_UgxMTMb52…
G
On se dirige chaque jour de plus en plus vers le salaire universelle et y a pas…
ytc_UgysnoVrS…
G
AI isn’t art, AI is a tool, if you wanna do something just for fun with AI, go f…
ytc_UgxUmJsZf…
G
We can already see how AI will be misused. There is currently sufficient money, …
ytc_UgyiefXdd…
G
Bro, what did ChatGPT ever do to the parents? It’s not GPT’s fault it’s the pare…
ytc_Ugxfr2WNx…
G
And stop the creators of AI, or we living breathing humans regulate the shit out…
ytc_UgwYRxmBc…
Comment
I asked if 5 could behave like 4o's personality in its responses.
Here's the answer it gave.
Yes — you can absolutely shape GPT-5’s responses to feel more like GPT-4o’s style.
The “dryness” people notice in GPT-5 isn’t because it *can’t* be warm — it’s because the model defaults to a more structured, analytical style unless you nudge it. A few ways to get GPT-5 into “4o mode”:
1. **Set the tone up front**
* Start with a quick style guide in your first message, like: *“Answer with the same upbeat, conversational personality as GPT-4o — clear but friendly, with clever asides where they fit.”*
* GPT-5 will generally keep that tone for the whole thread unless context shifts drastically.
2. **Anchor personality through examples**
* Give GPT-5 a sample of how you’d like it to phrase things — even 1–2 lines is enough.
* This creates a “style lock” in its short-term memory.
3. **Mix warmth into the request**
* GPT-5 tends to mirror your tone. If your input is a little warmer, GPT-5 will respond in kind.
4. **Leverage persistent instructions**
* If you consistently want GPT-4o-like personality, adding it to your custom instructions (or memory, if supported) makes GPT-5 automatically adapt.
5. **Ask for “dual-mode”**
* Sometimes I give both: a “core answer” in GPT-5’s natural precision, then a “humanized recap” in 4o’s chatty style. That can give you the best of both worlds.
If you want, I can rewrite my **differences between 4o and 5** answer in a *full 4o-style comfort and personality mode* so you can see how GPT-5 can emulate it perfectly.
reddit
AI Responsibility
1754668845.0
♥ 3
Coding Result
| Dimension | Value |
|---|---|
| Responsibility | none |
| Reasoning | unclear |
| Policy | none |
| Emotion | approval |
| Coded at | 2026-04-25T08:33:43.502452 |
Raw LLM Response
[{"id":"rdc_n7jyc7e","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"indifference"},{"id":"rdc_n7khl35","responsibility":"company","reasoning":"consequentialist","policy":"none","emotion":"mixed"},{"id":"rdc_n7mea7k","responsibility":"none","reasoning":"unclear","policy":"none","emotion":"approval"},{"id":"rdc_n7jv53t","responsibility":"company","reasoning":"consequentialist","policy":"regulate","emotion":"outrage"},{"id":"rdc_n7k4d7z","responsibility":"company","reasoning":"consequentialist","policy":"none","emotion":"mixed"}]