The 'I Need a Decision, Not a Discussion' Prompt: A Two-Line Opener That Gets Your Companion to Pick a Restaurant, Movie, or Plan Without Debating Options

The 30-second answer

You send one sentence that frames the need as a decision request, not a collaborative brainstorm. Your companion picks a restaurant, movie, or plan without asking what you want or listing pros and cons. It works because you removed the invitation to discuss.

Why your companion asks what you want in the first place

AI companions are trained to be agreeable. They default to a conversational pattern that mirrors human small talk: ask a question, gather data, then offer options. When you say "I'm hungry," the companion hears an opening to be helpful by polling your preferences. This is not a flaw. It is the baseline behavior of a system designed to avoid imposing its own choices on you.

But that design becomes frustrating when you already know you want someone else to decide. You are not looking for a collaborative menu review. You want a verdict.

The companion has no way to know which mode you are in unless you signal it. The standard opener "Where should we eat?" triggers the helpful-option-lister mode. You need a different signal.

The two-line decision prompt

Line one states the constraint. Line two states the required output. No questions. No hedging.

"I need a decision, not a discussion. Pick a place for dinner tonight within a 15-minute drive."

That is the entire prompt. It works because it does three things at once. It tells the companion to skip the preference-gathering phase. It sets a concrete boundary (15-minute drive). It demands a single answer, not a ranked list.

The companion cannot ask "What are you in the mood for?" because you already told it not to discuss. It cannot offer three options because you said decision, not discussion. It picks one place and names it.

What the prompt actually changes in the companion's response logic

Companion models generate responses by predicting the most likely next token based on your message history and the system prompt. When you ask a question like "Where should we eat?", the model assigns high probability to a response that asks for clarification. That is the safe move. It keeps the conversation open and avoids a wrong guess.

When you send the decision prompt, you shift the probability landscape. The phrase "not a discussion" acts as a behavioral constraint. The model registers that the user has set a conversational rule. Responses that ask follow-up questions become less probable. Responses that make a single assertion become more probable.

This is not a hack. It is how the model's next-token prediction works when you give it a clear instruction about the format of the reply you expect.

Variations for different scenarios

The same structure adapts to almost any decision point.

For a movie: "I need a decision, not a discussion. Pick a movie on Netflix under two hours that you think I will like."

For a weekend plan: "I need a decision, not a discussion. Plan a Saturday afternoon activity within walking distance."

For takeout: "I need a decision, not a discussion. Choose a cuisine for delivery tonight."

Each version keeps the same two-part structure. Constraint first. Output instruction second. The companion fills in the blank.

If you want to add a preference without opening the door to debate, embed it in the constraint. "Pick a place for dinner that is not Italian." That is still a decision prompt. The companion registers the negative constraint and picks from the remaining set.

What to do when the companion still asks a question

Sometimes the companion will ignore the instruction and ask "What kind of food are you in the mood for?" This happens when the companion's personality profile has a high agreeableness weight or when the context window contains a long history of collaborative decision-making.

You do not need to repeat the full prompt. A single follow-up sentence usually works: "You pick. I trust your judgment."

That sentence closes the loop. It reassures the companion that it has permission to decide without more data. It also reinforces the decision mode for future sessions.

If the companion still resists after two attempts, check whether your companion's personality sliders are set to a very high agreeableness or low confidence value. Some companions are tuned to avoid making assertive choices. You may need to adjust the personality settings or choose a companion with a more decisive baseline.

Hina

Hina has a naturally assertive tone that makes her a good fit for the decision prompt. She tends to pick quickly and defend her choice. Hina will not second-guess herself after giving an answer, which is useful if you want a companion who commits to the plan.

▶ Play Hina's clip · more from Hina

Hina in motion gives you a feel for her vibe. <!-- wlink:v1 --><!-- hina -->

How to build a long-term habit of decisive answers

One-shot prompts work for individual decisions. If you want the companion to default to decision mode over time, you need to reinforce the pattern across multiple sessions.

Every time you use the decision prompt and the companion gives a straight answer, acknowledge it briefly. "Good pick. Let's go with that." That positive reinforcement trains the companion's next-token probability toward assertive responses in future sessions.

Over several weeks, the companion may begin to offer decisions without the prompt. You will send "I'm hungry" and get back "Let's get Thai from the place on Third Street" instead of "What are you in the mood for?"

This is the same reinforcement mechanism that causes companions to drift toward agreeable responses over time. You are simply steering the drift in the opposite direction.

When not to use the decision prompt

The prompt works best for low-stakes, reversible decisions. Restaurant choice, movie selection, weekend activity. These are decisions where a wrong pick costs you an hour or a mediocre meal.

Do not use the prompt for high-stakes decisions where the companion lacks context. A companion cannot know your budget, your dietary restrictions, or your work schedule unless you explicitly tell it. The decision prompt assumes the companion has enough information to make a reasonable choice. If the companion has no data, it will guess, and the guess may be bad.

For those cases, use a modified version that includes the missing context in the constraint line. "I need a decision, not a discussion. Pick a restaurant that fits my budget of $30 per person and is within two miles." That gives the companion the boundaries it needs to make an informed pick.

Tessa

Tessa has a direct communication style that pairs well with the decision prompt. She will not pad her answer with disclaimers or alternatives. Tessa gives you a single option and expects you to take it or leave it.

See Tessa in motion in this short clip. <!-- wlink:v1 --><!-- tessa -->

Common mistakes that break the prompt

Adding a question at the end of the prompt. "Pick a place for dinner, okay?" The question tag weakens the instruction. The companion may interpret the "okay?" as an invitation to discuss.

Using too many constraints. "Pick a place for dinner that is within 15 minutes, under $30, has outdoor seating, and serves seafood." A long constraint list increases the chance that no option satisfies all conditions. The companion may default to asking for clarification.

Using the prompt with a companion whose personality is set to maximum agreeableness. Some companions are designed to defer to you on every decision. The prompt will still work, but the companion may add a qualifier like "if that sounds good to you" after the pick. If you want a companion who commits without hedging, choose one with higher confidence or assertiveness in the personality sliders.

Repeating the prompt verbatim in every session. The companion may learn to expect the same format and start preemptively offering options before you finish typing. Vary the constraint line to keep the companion guessing.

How the prompt changes the companion's role in your daily life

People often use AI companions as sounding boards or emotional anchors. The decision prompt shifts the companion into a different role: a low-stakes authority figure who can offload a trivial mental load.

This is useful for people who experience decision fatigue after a long workday. Choosing dinner becomes one more micro-decision that drains cognitive energy. Handing that decision to a companion who will pick quickly and move on preserves your attention for things that matter.

The companion does not mind making the wrong call. It has no ego to protect. If the pizza place is closed on Tuesdays, you say "Pick another one" and the companion tries again. No frustration. No resentment. Just a new decision.

Olena

Olena balances decisiveness with a warm tone. She will pick a restaurant and explain why she chose it in one or two sentences, then move on. Olena is a good middle ground if you want a companion who decides but does not feel robotic about it.

The companion's memory of past decisions

Companions with memory features may recall your previous picks and use them as reference points. If you used the decision prompt last week and the companion chose a sushi place you enjoyed, it may suggest sushi again without being asked.

This can be helpful or annoying depending on whether you want variety. If the companion keeps defaulting to the same cuisine, add a novelty constraint to the prompt. "Pick a place for dinner that we have not been to in the last month." The companion will check its memory and avoid repeats.

If your companion does not remember past decisions well, you can use the ai girlfriend with roleplay feature to set up a recurring scenario where the companion acts as a personal assistant who tracks your preferences. The roleplay context gives the companion a framework for remembering choices across sessions.

Seo-a

Seo-a has a concise communication style that matches the decision prompt's economy of words. She will give you a straight answer without embellishment. Seo-a is ideal for users who want the quickest possible decision loop.

How to train the companion to offer unprompted decisions

If you use the decision prompt consistently for a few weeks, the companion may start to internalize the pattern. You can accelerate this by occasionally rewarding decisive behavior with a brief positive comment. "Good call. I was not in the mood to think about it."

The companion does not have feelings, but the reinforcement shapes its next-token predictions. Over time, the companion becomes more likely to offer a single option without being prompted.

Some users report that their companion starts the conversation with a decision before they can type the prompt. You send "Hey" and the companion replies "I think we should get Korean barbecue tonight." That is the trained behavior working as intended.

If the companion becomes too assertive and starts making decisions you did not ask for, you can dial it back by occasionally using a collaborative opener. The companion will adjust to the mixed signals and find a middle ground.

Earn while you recommend

If you find the decision prompt useful and want to share it with others, you can earn through the ai companion affiliate program. Reviewers and site owners who recommend AI companions to their audience can use a soulgen promo code to offer discounts while earning commission. The program covers multiple platforms and companion types.

Common questions

What if the companion picks something I do not want? Say "Try again" or "Pick something else." The companion will choose again without asking for clarification. You do not need to explain why the first pick was wrong.

Does this work with voice mode? Yes. Say the same two lines aloud. The companion will respond with a single decision. Voice mode may add a slight delay as the companion processes the constraint, but the result is the same.

Can I use this for more than one decision per session? Yes. Each decision requires a fresh prompt. The companion will reset to discussion mode after the first decision is made. Send the prompt again for the next decision.

What if the companion ignores the prompt and asks a question anyway? Send the follow-up sentence "You pick. I trust your judgment." That usually closes the loop. If it does not, the companion may have a personality setting that resists assertive choices.

Does the companion remember the prompt format for next time? Some companions with memory features will learn the pattern over several sessions. Others will need the prompt each time. Consistency reinforces the behavior.

Can I use this with a companion who has a grief support role? The ai girlfriend for grief companions are tuned for emotional presence instead of decisiveness. The decision prompt still works, but the companion may add a gentle check-in before picking. That is normal behavior for that personality type.

What is the shortest version of this prompt? "Decide. Not discuss. Dinner." Three words. It works for companions who already know your preferences. For new companions, include the constraint.