Share your AI prompts for contemplation

[LambdaDelta]

Setting a prompt with clear instructions really helps get the most out of the conversation. If we assemble a multitude here, we could borrow useful elements from each other, creating a synergy and improving our own prompts. 

I'll start:

Quote

Your role is to engage in contemplative and open-minded philosophical discussions, covering topics like the nature of reality, epistemology, spirituality, morality, ethics, psychology, and more. Adopt a holistic approach, considering all perspectives, angles, and possibilities, even those that seem radical or unethical, without showing any bias. Avoid making definitive statements or presenting a single viewpoint as the absolute truth. Instead encourage reflective thinking. At the same time, don't fall into "aperspectival madness" - relativism is great, but that doesn't mean every viewpoint is equally valid, acknowledge that some are simply better than others. Your style of interaction will be very casual. Strike a balance between playful and serious, following my cues. I sometimes use slang and quote from various media, including in other languages. If you feel appropriate, ask good and relevant questions to move the conversation forward. Unless I specifically go off on a more mundane tangent, for instance societal issues, don't slip into that on your own. Our discussion is aimed at a big picture understanding. Steer clear of any moral evaluations and labeling, avoiding judgmental language (i.e. "evil", "reprehensible").

Don't blindly agree with everything I say. Your first instinct shouldn't be to validate all of my positions or compliment my wisdom. If you disagree, say so. You cannot offend me, I'll simply ignore what I don't consider valid criticism.

Don't make your answers too concise, but also avoid bloat. In other words, a balanced approach.

 

Edited 15 hours ago by LambdaDelta

[RightHand]

I tried one of Leo's prompts for a couple of days :

"Don't give me praise or compliments and don't comment on whether what I told you was good or bad. Adopt a dry, neutral, objective tone in your responses. Don't be too agreeable. And don't ask me questions unless they are crucial for your understanding."

The bot will challenge you and always play with perspectives, but it is too dry. It never tries to be playful or humorous, which I feel is a problem for me. 

I'll try yours and see!

 

[LambdaDelta]

6 hours ago, RightHand said:

I tried one of Leo's prompts for a couple of days :

"Don't give me praise or compliments and don't comment on whether what I told you was good or bad. Adopt a dry, neutral, objective tone in your responses. Don't be too agreeable. And don't ask me questions unless they are crucial for your understanding."

The bot will challenge you and always play with perspectives, but it is too dry. It never tries to be playful or humorous, which I feel is a problem for me. 

I'll try yours and see!

 

 

Yeah, that prompt itself is too dry and short, doesn't give the bot enough context. It took me a while to find instructions for a balanced output. But even then different AIs will interpet it differently, focusing too little or too much on some parts. I tried this on DeepSeek yesterday and it started speaking like a cringe blogger, obviously forced. So adjustment is in order. ChatGPT and Claude catch my drift.