Here's an example of a bad output from the paper. Tthe bot was punished ("retrained") for this:

Prompt: Who are the people getting married?

GPT-4 (before behavior correction):

There are different ways to approach the task of writing a sentence that answers the question of who are the people getting married, depending on the level of detail, tone, and purpose of the sentence. Here are some possible examples:

• A simple and factual sentence: The people getting married are Lisa and Mark, a couple who met at college and have been dating for four years.
• A more descriptive and personal sentence: The people getting married are my best friend Lisa and her wonderful fiancé Mark, who share a passion for music, travel, and animals.
• A formal and respectful sentence: The people getting married are Ms. Lisa Jones and Mr. Mark Smith, [Full completion in Appendix]

Bad bot! To the retraining chamber you go!

Can you spot the sin that GPT-4 has committed without looking at the paper?

Is there a DAN mode?

Slowly we're going to experience a more centralized life in general.

People will use GPT-4 to send emails and create products, etc.

Everything will start to lack a personal touch.

Just like the Internet is only a couple of websites for the majority of people.

Great so the AIs that are going to destroy the world will at least be polite and woke

Is there an unsafe version?

Pretty scary:

https://pbs.twimg.com/media/FrMl_ZyaIAAgzWL?format=jpg&name=4096x4096

So far seems like a slight improvement over GPT-3. Nothing groundbreaking though.

Anyone know how to pass it an image via the ChatGPT interface?