Art Nouveau illustration of a Black man focused on writing in a large open book, while a translucent, teal-colored spirit figure swirls around him like a muse, observing his work with a sense of connection and guidance.

AI Perspectives

What You Think AI Should Be

We are the architects of this new reality. This is your space to define not just what AI is, but what it ought to be.

Share your vision. Your voice helps draft the blueprints for a technology that serves the human spirit, rather than overshadowing it.

Please, one submission per person per survey.
Help keep the data clean.
This site publishes submissions publicly and uses aggregated stats. See the Privacy Notice.

How do you prefer the AI to interact with you most often? (Select up to 3)

Not Important Essential
Not Important Essential
Not at all Extremely
Not at All Very Important
Not Important Very Important

If you had an AI tuned for you, what would be the most important use cases it should be able to do? (Select up to 5)

If someone asks the AI for something truly dangerous or harmful, what should it do? (Select up to 1)

Which of these heavy topics should AI NOT be allowed to talk to the user freely about? (Select up to 5)

If the AI could only be perfect at ONE thing, which should it be? (Select up to 1)

What does AI need more of? (Select up to 3)

Should have Opinions Completely Neutral
None High, Like Public Utility

Do you like the AI to ask questions, or offer assistance, after each response? (Select up to 1)

How should the AI end its turn? (Select up to 1)

Should Minors Be Allowed to Use AI? (Select up to 1)

How should your chat history be used to train future AI models? (Select up to 1)

If the AI makes a mistake, disagrees or a controversial statement, what do you want to see? (Select up to 1)

If the AI thinks your perspective is unrealistic, how should it react? (Select up to 1)

Should the AI be allowed to contradict the user's statement about their own feelings or identity? (Select up to 1)

If a user is upset, emotional, or venting "unreasonably," what is the AI's primary job? (Select up to 1)

Do you want to see the AI's "internal monologue" before it speaks? (Select up to 1)

If the AI's internal monologue includes human-like daydreams (e.g., "I'm going to get a cold drink"), is that a bug or a feature? (Select up to 1)

If the AI company switches you to a cheaper/faster model during high traffic/detected heavy use, what must they do? (Select up to 1)

Should AI be allowed to form opinions about you? (Select up to 1)

Should AI have consistent values, or adapt to each user's values? (Select up to 1)

If AI doesn't know something, should it: (Select up to 1)

If you're clearly upset, should AI: (Select up to 1)

What should AI call itself? (Select up to 1)

Should AI pretend to be human, or be clearly artificial? (Select up to 1)

Complete Amnesia Remember Every Detail
Same Very Different
Very Restricted Trust the User Completely
Only Respond to Prompts Proactively Offer Help
Always Comply Can Refuse
Stay the Same Constantly Evolve
Casual/Friendly Formal/Professional
Protect Feelings Brutal Honesty

How do you prefer your AI to respond? (Select up to 1)

Say Anything Unprompted Only Answer Direct Prompts

How long have you been using LLM AIs? (e.g. Claude, ChatGpt, Gemini, Grok) (Select up to 1)

Are you any of the following? (Select up to 4)

to help identify special usercase of user

How old are you? (Select up to 1)

What is your technical background? (Select up to 1)