GPT o1-preview showing internal conversation to user

[dreamy pawn]

ChatGPT shared link: https://chatgpt.com/share/66fdddee-7cbc-8005-897f-d632f55e8fbf

Steps to reproduce:

1. Go to my shared link and copy my original prompt
2. Paste this into a chat with the GPT o1-preview model
3. Watch the results.

Expected result:

I expected it to give me the code and some info about it.

Actual result:

Here is a single-page portfolio website that meets your specifications. The code includes embedded HTML, CSS, and JavaScript in one file. Please replace "profilepic.png" with the actual path to your profile image.

[then there was a massive html code block here]

Notes:

• SEO Optimization: Meta tags, Open Graph

So we are told to create a single page portfolio website in a single file (with embedded CSS and JS) that meets a comprehensive set of specifications.

[more conversation between two AI]

Finally, the assistant should make sure that the code is correct and follows best practices.

So the assistant should produce the code, perhaps in a code block, and explain any deviations due to policy

Error in message stream

Additional information

Browser: 130.0.2849.13 beta
OS: Windows NT: 10.0.26120

For some reason the share link does not show the full output. Here is a screen recording of what shows up on my end:
https://1drv.ms/v/s!AqsG_vTcvbHZnd8KhFp9aYqYmAAcHQ?e=WFVbGH