I’ve lost count of how many times I’ve watched an AI assistant turn a questionable startup idea into a convincing plan.
Ask it about an idea and it can give you a market, a business model, pricing, features, competitors, and a roadmap.
The problem is that most of the time, you’re still talking to something that is trying to answer your question.
So we built Clarno around a different question:
What if the system’s job was to argue with the idea instead?
Clarno starts by reading the real market: who is already struggling, what they use today, what they pay, and what evidence actually exists.
Then the idea goes into a simulation with market agents.
These aren’t static persona cards. Agents have personas, budgets, memory, and their own reasons for pushing back. They remember what happened in earlier rounds, challenge the offer, and update their position as the idea changes.
The point isn’t to make the simulation sound realistic.
The point is to make the founder’s assumptions harder to hide.
The workflow
Clarno currently runs through six beats:
Scan → Argue → Verdict → Design → Case → Launch
The first three make up The Test.
Scan reads the market first.
Argue puts the idea in front of the cohort.
Verdict gives you a decision with the evidence behind it.
If the answer is FAILED, the run doesn’t simply stop.
Clarno maps the opportunity into three possible directions, with the strengths and weaknesses of each. The founder chooses which direction to pursue, and the market agents pressure-test that choice.
The important part is that Clarno doesn’t pretend every number is known.
Our rule is simple:
No source, no number. Not found is a finding.
If we can’t find the evidence for a figure, we don’t manufacture one to make the verdict look complete.
Why build another AI product?
Because the hard part of starting a company usually isn’t generating more ideas.
It’s knowing which idea deserves more time.
A chatbot can help you build almost anything.
We wanted to build something that could make you stop before you build the wrong thing.
Clarno 2.0 is now live. The first scan is free, with no credit card required.
Try it: https://www.clarno.ai/
I’d genuinely like to hear from other builders:
What’s the strongest reason you’ve ever had to kill an idea?
Top comments (0)