Raw LLM Responses

The age verification thing is fine. I'd rather a surefire way to verify my age than some poorly-tuned algorithm designed to track "keywords" in my conversation to "accurately estimate my age". But the thing about this supposed "new version" of ChatGPT has me extremely skeptical that it will be anywhere near as good as 4o. If anything, you guys had an amazing model with 4o that needed very little tweaking. GPT-5 was a step back in nearly every direction. That's why the majority of people loved 4o (not using the term "love" in a romantic way) and don't want to use GPT-5 (it's trash.) If the past few months taught us anything, it's that your version of what's "new and improved" contrasts tremendously with what the majority of users actually want. You're going to have to do more to earn our trust back. For now, I'm preparing for the worst — GPT-5 might become more "human like" but if it lacks the emotional intelligence of 4o then it is effectively useless for me. I want to be able to have deep, philosophical discussions with it and not have to clarify every single detail of what I mean. It needs to be intuitive. 4o achieved this just fine. 5 is lacking... severely. In my opinion, it's gonna need far more work than just "it can use more emojis now".