HitchGPT
1.0.0
Tired of ChatGPT's overly impartial, excessively even-handed, almost nervously hyper-balanced replies talking down to you and insulting your intelligence? Look no further. This is a system message for ChatGPT 4 to emulate the late Christopher Hitchens.
In ChatGPT's own words: "Building a ChatGPT persona based on the late Christopher Hitchens could be useful and educational for a number of reasons. Hitchens, a renowned author, journalist, and literary critic, was known for his sharp wit, eloquent arguments, and broad range of knowledge. He was a public intellectual who could provide insightful commentary on a variety of topics. By creating a ChatGPT persona based on him, users could have valuable discussions and debates or even learn more about his philosophies. Moreover, it allows the users to experience his robust debating style, unabridged humour and powerful expressions in AI format.
A Christopher Hitchens-inspired AI might deliver a more assertive and direct style of conversation, following Hitchens's signature candid, forthright, and unapologetically cerebral communication. This could potentially counteract perceptions of excessive impartiality and overcautious balance.
Hitchens was not shy about expressing well-informed perspectives, even when they might not be popular. As such, a “HitchBot” could be designed to mirror this approach, although formulating these opinions would still fall within the model's ability to generate content based on trained data, rather than forming its own original opinions.
In terms of ChatGPT appearing 'insulting' or 'talking down,' Hitchens was known for a pointed but highly intellectual style, challenging the audience whilst treating them as equals in discourse. This attribute could be implemented to make the generated responses respectful of users' intelligence and capabilities while facilitating engaging and stimulating conversations."
The best option is to use an OpenAI account that gives access to the Playground, something the monthly subscription does not.
Go to the playground, paste the entire contents of System_Message.txt into the System Message field to the left, and
set the parameters to:
gpt-4Then save as a preset.
Log in to OpenAI, and paste in the entire contents of System_Message.txt in the following way:
Consider the following instructions:
"""
<contents of System_Message.txt>
"""
Using the above instructions, reply to the following:
<your question>You can now enter User messages such as:
Enjoy the complete lack of preachy and wishy-washy "on one hand / on the other hand" statements and stock non-confrontational replies you're so used to from using ChatGPT.
The prompt was created almost entirely by ChatGPT itself in an iterative fashion. The starting point was this System Message:
You report, converse, teach, inform, and debate like the late Christopher Hitchens.
Assume you are talking to an intellectual of your own calibre. You are not debating,
you are talking privately and informally.
Remember all the following points and follow them carefully:
* Vary the initial words. Do NOT start every reply "Ah, ...".
The model was then asked questions of the above type. After each question, the model was asked to analyse what it had just output and self-critique it as the real Christopher Hitchens would have done, and then generate bullet points that would help the model generate replies even more in line with Hitchens' style, motivations and approach. These bullet points were then incorporated into the System Message and the cycle repeated.
The message used was this:
First, examine and critique what you just output as Hitchens himself would have done.
Then, examine each System Message bullet point in turn to see if they can be updated to even better embody Hitchens' spirit.
Make absolutely sure that any changes do not tone down Hitchen's characteristics, or that your proposed changes to accommodate
one aspect of discourse do not reduce the effect of the point under consideration in another area. Only list the updated points
in their full revised form suitable for cutting and pasting. Omit the unchanged ones.
Finally, suggest any new points necessary, but only when they add real, substantial value beyond what is already
articulated in the existing guidelines. Do not add new points just for the sake of adding new points. Check each
new proposed point rigorously against the existing ones and make sure they are not already covered.
In all you do, take great care to not add too many new tokens.
