Your AI chat can become a deliberation partner. When you are weighing a decision, revising a draft, or developing an idea, it is easy to ask AI for agreement. A more useful conversation makes assumptions visible, gives reasons for objections that matter, and revisits recommendations when new information arrives.
At SOSYU, we developed the Istisarelist Mode prompt from our experience deliberating with AI and examining responses through follow-up questions in the laboratory. It is designed for anyone who wants thoughtful feedback while keeping responsibility for the decision.
WHAT IS ISTISARELIST MODE?
Istisarelist Mode is a copyable starter prompt that tells an AI assistant how you want to work together. It brings three Istisarelism principles into everyday conversations: reasoned objection, diminishing objection, and reasoned closure. It asks the assistant to understand the goal, state uncertainty, and raise concerns in proportion to their effect on the decision.
If a project proposal rests on three assumptions, for example, a useful deliberation partner identifies which assumption could change the decision. It focuses on the information that determines the first step. When you supply that information, it revisits its recommendation.
COPYABLE ISTISARELIST MODE PROMPT
“Work with me as a deliberation partner throughout this conversation. First, understand my goal and the criteria for the decision. If missing information could materially change the outcome, ask a short, focused question. If you can proceed under a reasonable assumption, state it.
Distinguish what you have verified, what you infer, and what you cannot verify. If you have not examined a source, say so. When an important fact is missing, explain what we should check.
If you agree with a proposal, explain why. If you see an error, contradiction, or risk that could affect the decision, raise a concise objection and give your reasons. Where possible, offer an alternative or a small, reversible step. Once I understand an objection, repeat it only when new information or a significant risk warrants it.
Revisit your recommendation when new evidence or a stronger reason appears. If you change your view, explain what changed. If there is too little information for a decision, state what remains unresolved. When useful, end with a short recommendation and a next step. The final decision is mine.”
HOW TO USE IT WITH CHATGPT, CLAUDE, OR GEMINI
1. Copy the starter prompt inside the quotation marks.
2. Open a new conversation in your preferred AI chat tool and paste the prompt as your first message.
3. Send the question, decision, or text you want to discuss. Include your goal and any limits on time, budget, or available resources.
4. When new information arrives, share it and ask, “Does this change your earlier recommendation?”
This is a user prompt entered into a chat. It does not change the platform’s system settings. Models may follow it differently, and their answers still need review.
TRY IT YOURSELF WITH TWO SAMPLE QUESTIONS
Decision example: “I want to start a new project this week. I have two evenings available and have yet to confirm that the idea meets a real need. What small step should I take first?”
Writing example: “My announcement says, ‘This method saves everyone time.’ I only have my own experience to go on. Rewrite the claim to match my observation and tell me what I could measure to make a stronger claim.”
Ask the same question in two fresh chats and use the starter prompt in one of them. Check whether each response names the assumption that affects the decision, explains any objection and proposed next step, and explains any change in recommendation after you share new information. One trial can help you see what is useful for your own work.
FREQUENTLY ASKED QUESTIONS
Will this AI prompt prevent mistakes?
The prompt asks the assistant to distinguish information from inference and to state what it cannot verify. It cannot guarantee accuracy. Verify any numbers, quotations, or sources that matter to your decision.
Will the AI object to everything I propose?
The prompt calls for reasoned objections when a concern could affect the decision. Once the concern is understood, the assistant should move on unless new information or a significant risk changes the situation.
When should I try Istisarelist Mode?
Try it before publishing a draft, while evaluating a project idea, when choosing between options, or when you suspect an early assumption is driving your decision.
Give it a try with a question you are actually working through. See which assumption the AI notices and whether new information changes its recommendation.
More about Istisarelism:
https://sosyu.com/en/istisarelizm
TRY ISTISARELIST MODE IN YOUR OWN CHAT
Visit the SOSYU Experiments section to copy the Istisarelist Mode prompt in one click. Paste it into a new chat in your preferred AI tool. Then ask your own question, discuss a decision, or share a draft. Consider what changed compared with similar answers you received before.
If you like, rate how much the prompt helped your work on a scale of 1 to 5. You can see the overall rating and the number of responses on the same page. Your feedback helps us improve the prompt.
Try Istisarelist Mode and share your experience:
https://sosyu.com/en/experiment/1
News
Turn Your AI Chat into a Deliberation Partner

Tags: Istisarelist Mode,Istisarelism,AI chat prompt,ChatGPT,Claude,Gemini,human AI interaction,reasoned objection
