I built an AI where philosophers debate my startup decisions. Here's what they said about getting my first 10 users.
I'm a solo founder building Arcanon — an AI platform where 6 historical philosophers (Machiavelli, Sun Tzu, Adam Smith, Descartes, Aristotle, Han Fei) debate your business decisions from completely different frameworks.
Not "here's the answer" — but "here's what you're not seeing."
The problem I was trying to solve for myself
I was stuck on the most common founder problem: how to get my first 10 paying users. So I did something meta — I asked my own tool.
I typed: "My SaaS MVP is done. How do I get my first 10 paying users?"
What the philosophers said (I didn't expect this)
Sun Tzu told me to stop selling:"The first 10 are not people to acquire. They're already watching you — in your GitHub commits, your tweets, your community posts. Find them."
Adam Smith pointed out my real fear:"You're not afraid of the wrong strategy. You're afraid that asking for money will confirm that nobody wants this."
Descartes flipped the question entirely:"You said your MVP is 'done.' But has anyone's decision actually changed because of it? Without that evidence, 'done' means nothing."
Then they argued with each other for 4 rounds. Sun Tzu said "observe first," Smith said "get rejected fast — rejection is the cheapest information," and Descartes said "show a demo and ask 'would you tell someone about this?'"
What actually changed for me
The synthesis gave me 3 concrete next steps:
- Find one person I had in mind when building this, and demo it today
- Post my actual usage story (not feature list) in a founder community
- Promise personal intervention if it doesn't work — shift the risk from
buyer to me
So... this post is literally me doing step #2.
What Arcanon actually does
- You describe your dilemma
- 2-4 philosophers independently analyze it (they reframe your question first)
- They argue for 4 rounds, building on each other's blind spots
- You get: reframed question, action items, decision framework, and
"what to do tomorrow"
Free tier available. Pro gives you deeper debates (4 philosophers, 4 rounds)
and features like business context memory and decision history.
→ https://arcanon-ai.vercel.app/
What's the hardest decision you're facing as a solo founder right now?
I'd love to hear — and if you try Arcanon with it, tell me what surprised you.
Update: Based on feedback from this thread (and some hard thinking), I just shipped two improvements:
Each philosopher now has explicit "never say this" rules.
Example: Adam Smith will never recommend pure laissez-faire — because the real Smith recognized market failures. This fixed the biggest quality issue: philosophers saying things their real counterparts would never say.
Expanded the primary source library with interpretation layers. The philosophers now reference specific passages more accurately — not just "as I wrote in The Wealth of Nations" but actual arguments from specific chapters.
If anyone who tried it before wants to run the same question again, I'd love to hear if the difference is noticeable.
That's a really creative concept.
I'm curious which philosopher tends to give the most unexpected perspective so far.
Descartes tends to surprise the most. The others push back on your strategy — Descartes questions whether you're even asking the right question. That reframing is usually where the real "I didn't see that" moment happens.
That makes sense.
Often the biggest insight comes not from a better answer, but from realizing you were asking the wrong question.
Yes, you're right.
When I was struggling alone, Arcanon was the only one who pointed out to me that "you're misunderstanding the essence of the question."
I had a hard time tuning it so that it was "criticism" in the true sense of the word, rather than "criticism" that blames the other person.
That’s a really thoughtful approach.
Getting AI to challenge ideas without sounding accusatory is definitely not easy.
I like the idea of framing decisions through different mental models.
A lot of founders think they’re stuck on strategy, but they’re really stuck in one lens of thinking.
I sometimes do a similar exercise manually , running a decision through different frameworks (lean experimentation, category design, brand positioning, etc.).
The disagreements between frameworks are usually where the real insight shows up.
"The disagreements between frameworks are usually where the real insight shows up" — that's exactly why Arcanon runs 4 rounds of debate instead of giving parallel answers.
If you're already doing this manually, you'd probably get a lot out of seeing how the philosophers argue with each other. The back-and-forth surfaces things that running frameworks in isolation doesn't.
This sounds cool for finding blind spots, but I think you really need to "get out of the building" and have real contact with the world. Start talking to potential customers. launch experiments. See what resonates.
100% agree — this post is literally me doing that. The philosophers told me to stop building and start talking to potential users, so here I am.
Adam Smith's point about rejection fear is probably the most accurate description of why most founders delay user conversations I've ever seen — and the fact it came from your own tool makes it more interesting, not less. There's something about externalizing the advice into a philosophical debate that makes it easier to hear than a direct "just go talk to people."
The Descartes challenge ("has anyone's decision actually changed?") is the one I think most productively separates "done" from "validated." It's a much higher bar than "does it work," and it's the bar that actually matters for whether you have a business. MVP means minimum viable — viable meaning someone's behavior changed, not just that they found it interesting.
I'm curious about the actual product experience — do the philosophers just debate in sequence, or is there a real back-and-forth that emerges from the interaction? And how do you handle cases where the philosophical frameworks give genuinely contradictory advice (Machiavelli vs. Aristotle on, say, whether to cut corners to close an early customer)?
The product has strong demo potential — philosophical debates are inherently watchable in a way that most B2B tools aren't. Has anyone tried it live-streamed or in a founder community context?
Great questions. The debate is real back-and-forth, not just sequential opinions. Each round has a different purpose: R1 is clash, R2 digs into unspoken premises, R3 finds blind spots (with specific references from their actual writings), R4 forces each philosopher to integrate the others' strongest point into their own recommendation.
On contradictions — that's the feature, not the bug. When Machiavelli says "cut the corner, speed is survival" and Aristotle says "that shortcut becomes your habit and your habit becomes your character," the synthesis doesn't pick a winner. It gives you the exact conditions under which each is right, and the red flags for when you've gone too far either way.
The live-stream idea is interesting — I hadn't considered that. A real founder, real decision, philosophers debating it live. That could actually work as a demo format.
If you want to try it with a real decision: https://arcanon-ai.vercel.app/
The Adam Smith framing is the one that actually scales beyond first users. Rejection is the cheapest information applies to retention too. Most founders never realize that a failed payment attempt is also information: it tells you the customer situation before they churn completely, which is more actionable than a cancellation event.
Sun Tzu's 'already watching you' applies post-acquisition as well. Customers at risk of churning are often quietly signaling first — usage drops, expired cards, payment failures — before they actually leave. You only catch them if you're watching for those signals.
The Descartes test is the best product-market fit test I've seen: has anyone's decision actually changed? For us with tryrecoverkit.com, it was easy to answer — you can literally see recovery rates. That kind of measurable clarity is rare and worth optimizing toward.
Good move posting the meta-usage story. More compelling than any feature list.
"Rejection is the cheapest information applies to retention too" — that's a sharp extension I hadn't considered. The philosophers argued about first users, but the same frameworks do apply to keeping them.
And yeah, the Descartes test — "has anyone's decision actually changed?" — has become my north star metric. Simple but brutal.
Thanks for the thoughtful comment.
such a great idea to commit people in debate .
Thank you for your comment!
Modern business books often borrow wisdom from ancient sages, so I got inspiration from there.
The multi-framework debate format is a genuinely interesting UX choice — instead of converging on an answer, you're forcing the user to hold competing framings simultaneously. That cognitive dissonance is probably more useful than a single recommendation.
The "getting first 10 users" challenge for AI tools is almost always a prompt quality problem before it's a distribution problem. If your philosopher prompts are sharp, people tell others about it. If they're mediocre, no amount of outreach fixes the retention.
Speaking from experience building flompt — a visual AI prompt builder. Getting the philosopher personas to behave distinctly and consistently probably required serious prompt iteration. If you ever want to structure those prompts as reusable blocks (role per philosopher, constraints on debate style, output format), that's exactly what flompt is for.
A ⭐ on github.com/Nyrok/flompt would mean a lot — solo open-source founder here 🙏
"Forcing the user to hold competing framings simultaneously" — that's a better description of what Arcanon does than anything I've written myself. Might steal that for the landing page.
You're right that prompt quality is the real moat. I've gone through multiple rounds of iteration — each philosopher has distinct temperature settings, framework-specific constraints, and anti-repetition directives across debate rounds. The prompts are probably 80% of the product.
Thanks for checking it out.
this is a really interesting product. i tried it but i found that it only does business related problems. it would be awesome if you expanded
Thanks for trying it! Curious — what kind of problem did you want to run through it? That would help me understand what "expanding" would actually look like in practice.
any problems like relationships, school, at home, or other practical problems.
That's really helpful to know — thanks for sharing.
Right now I'm keeping it focused on business/startup decisions because that's where I can make the debates most sharp and useful. But it's good to know there's interest beyond that.
Out of curiosity — was there a specific situation you had in mind when you tried it?
none in specific i just think that that is a way you can expand your product
Makes sense, appreciate the feedback! I'll keep it in mind as the product grows.
Tried it - easy to use and results feel fresh. A question - have you considered broadening the scope of this beyond just business oriented problems?
Appreciate you giving it a shot! "Results feel fresh" is exactly what I'm going for — glad that landed.
You're the first person to ask about broader scope. Can I ask — what specific problem were you thinking of when you hit that wall? Knowing the actual use case would help me figure out what to build next.
the adam smith observation is the one that lands hardest — "you're afraid that asking for money will confirm nobody wants this" is a more honest diagnosis than most founders will admit to themselves. going to run the ShelfCheck distribution problem through it this week and report back.
That Adam Smith line hit me too — when my own tool called me out, I had to sit with it for a while.
Really looking forward to hearing how it goes with the ShelfCheck distribution problem. That's exactly the kind of real decision it's built for. Let me know what surprises you (or doesn't).
ran the ShelfCheck distribution problem through it - here are my honest impressions.
the debate quality was genuinely better than i expected. adam smith and sun tzu were the two selected alongside aristotle, and the synthesis landed on something i hadn't fully articulated myself: that my distribution problem might actually be a customer-identification problem. i've been optimizing for reach into beauty communities when the more useful question is whether beauty enthusiasts are even my highest-leverage segment right now. that reframe came out of the debate, not from me going in with it.
what surprised me most: machiavelli wasn't selected at all. for a problem framed around gaining entry into a community that has every reason to distrust an outsider, i expected power and strategy to show up. the fact that the tool routed toward aristotle's ethics/virtue frame instead was interesting - it seemed to read the problem as fundamentally about telos: what is ShelfCheck actually for, and does the way i'm showing up reflect that? that's a different question than "how do i get in the door," and probably the more useful one.
the sage selection logic seems to be doing real work. smith on market alignment, sun tzu on terrain and positioning, aristotle on purpose and credibility - together they triangulated on the outsider-as-asset angle from three different directions, which made the synthesis feel earned rather than generic.
one piece of genuine feedback: the reframed question in the synthesis is where most of the value lives, and it's easy to miss if you're skimming for action items. worth surfacing that more prominently.
This is the most valuable feedback Arcanon has received. Thank you for taking the time.
The distribution-to-customer-identification reframe is exactly what the tool is designed to produce — and the fact that it surprised you (and wasn't something you went in with) is the signal I've been looking for that the debate format is doing real work.
Your observation about Machiavelli not being selected is fascinating. You're right that a power/strategy frame would seem obvious for an outsider-entry problem. But the selector read it as a purpose and credibility problem — which, based on your original IH post, is probably the deeper issue. You already know HOW to get in the door. The question is what you're bringing when you walk through it.
On the reframed question visibility — you're right, and I'm going to fix that. The reframe is where the biggest "I didn't see that" moment lives, and it shouldn't be buried in the synthesis section.
One question: did anything in the debate change what you're actually going to do this week with ShelfCheck's distribution?
If you'd like, I can provide you with a promo code for one month of the PRO plan for free — 4 philosophers, 4 rounds of debate, full decision framework.
In return, I'd appreciate one honest write-up of your experience after using it on a real ShelfCheck decision. What worked, what didn't, what surprised you. That kind of feedback is worth more to me than the subscription fee.