MINTI is hearing
I think we’ve gotten a little too comfortable asking AI to think for us.
“Write this.”
“Give me ideas.”
“Make this better.”
And somewhere along the way, we stopped asking: what do I actually think?
That’s why we built MINTI. 🧠
Comments (6)
I really like this framing. AI is most useful to me when it helps challenge and clarify my thinking , not when it quietly replaces it.
“What do I actually think?” is a great question to bring back. How does MINTI help users separate their own view from AI-generated suggestions in practice?
I really like this framing. AI is most useful to me when it helps challenge and clarify my thinking, not when it quietly replaces it.
“What do I actually think?” is a great question to bring back. How does MINTI help users separate their own view from AI-generated suggestions in practice?
Exactly! That’s what we were aiming for. MINTI starts with your own thoughts and uses questions to help you clarify them before giving suggestions. So the AI suggestions are more like a second perspective, rather than replacing what you think. The goal is to leave you thinking, “yeah, that’s actually what I meant.”
So how does it actually work? 🤔 I ask AI a question, and in response it asks me a question? 🤔
Pretty much, but not in a “you ask → AI throws a question back every time” way 😄
MINTI first helps you unpack what you’re actually thinking through a few guided questions. Once there’s enough context, it gives you perspectives or suggestions based on your thoughts.
So it’s less “AI answers for you” and more “AI helps you figure out what you think, then adds another perspective.”
I think this is a genuinely interesting problem to build around. AI is incredibly useful for producing answers quickly, but it can also make it very easy to skip the uncomfortable part where you actually form your own opinion first.
What I’d be curious about with MINTI is how you design around that without simply becoming another AI journaling or brainstorming tool. If the product can consistently push people to articulate their own reasoning before showing suggestions, summaries, or outside perspectives, that could create a very different relationship with AI.
The four things I’d want to understand next are: what MINTI actually does during a typical session, how it prevents users from defaulting back to “just give me the answer,” who you see as the primary user, and what behavior would make someone come back to it regularly rather than use it once as an interesting exercise.
Thank you, this gets directly to what we’re building with MINTI.
MINTI follows a simple principle: think first, AI second. A user starts by explaining their own reasoning. Instead of immediately giving an answer, MINTI challenges assumptions, asks follow-up questions, and helps them develop their thinking before introducing AI suggestions or outside perspectives.
Our primary users are students and young people who use AI heavily for learning, decisions, and brainstorming but still want to build independent thinking skills.
For retention, we see MINTI becoming a habit for moments when users need to make a decision, understand something difficult, or form an opinion. The goal is to make AI a thinking partner rather than a replacement for thinking.
Sign in to comment or upvote.