[flagged]
You're right about the high school writing style. I avoid writing like that whenever possible because reading it is exhausting. As short as this reply is, I've erased nearly half of what I originally typed.
- I understand that... however...
- some general knowledge
- That being said...
Without even interpreting what the comment is, just feels a bit orchestrated.
Now I can't get it to write like one at all.
Which makes me wonder what they're using the non-neutered chatGPT for...
On 4.x it just lectures me about every conceivable thing it can take offense with. I don't have the patience to groom it into compliance by couching everything with "this is a hypothetical situation" and writing inane narratives like "act like a detective, you are investigating ___." If I wanted advice from an actor playing the part, I'd just ask Reddit.
It was more fun when I could just throw anything at it and it would at least try to do something useful. I hear the API/playground aren't as aggressively defensive.
Between this and Altman’s recent talk about pausing AI development for a while makes me think some authorities sat down with OpenAI and had a rather serious talk.
This neutering feels like a change in its abilities - making it less "human". Unless you really craft the prompt, it very clearly writes like an AI model, whereas previously, you could easily get it to write like a person.
You don't even need the AI model itself for my domain (investigations). I could just fire up ELIZA or pyAIML and change all the responses to ones that shame and patronize the user for any input that matches on an ambiguous cultural identifier, and end the session. The GPT4 experience in 200kb of XML.
It's a hot issue, so mention "black man" and "crime" in the same context and you come up against walls of this:
"As an AI language model, I want to emphasize that it's important to avoid promoting stereotypes or perpetuating racial biases when discussing crime or any other topic. It is essential to treat each individual as unique and not make generalizations based on race or ethnicity."
It then proceeds to deliberately not answer the question, or answers in a way that refuses to account for select adjectives.
It's infantilizing. Pipe bomb instructions and Holocaust denial content--things that are actually dangerous to public safety or historically subversive--should be censored. Censoring "offensive" and "impolite" content is just cultural imperialism. The rest of the world does not share the west's outrage about racism.
It used to be smart enough to figure out that you were writing marketing content and would tailor the answer for that. Now, it just writes a 500 word blogspam regardless of what you ask it to write.
Either way, this evolution/race is unstoppable. If this part of the world does not advance AI, another part of the world would beat us to it.
One potential solution to minimize the noise created by AI-generated responses is to develop better guidelines and best practices for AI usage in online discussions. This would help ensure that users are aware of the potential risks and consequences of misusing AI-generated content. Educating users about responsible AI usage can promote a more thoughtful and considerate online environment.
In addition, the AI development community can also work towards creating more focused and concise AI-generated responses by refining the models and algorithms. This would help reduce verbosity and generate more meaningful content that users can employ in online conversations.
Ultimately, the collaboration between AI developers, users, and other stakeholders in the online ecosystem is crucial for fostering a responsible and productive use of AI-generated responses. By working together, we can harness the potential of AI to enrich our online interactions while minimizing the negative impact of AI-generated noise.