I originally just wanted to find a good app idea.
That sounds harmless.
It is not.
A “good app idea” is a bit like “a simple integration”, “a quick MVP” or “we just need to sort out deployment”. It starts as a small thought and often ends with three dashboards, an auth solution, a pricing model, a pitch deck and a feeling that you should probably open YouTube to get your brain back.
I wanted to avoid that trap.
So I used AI as a sparring partner.
Not just to spit out 20 ideas.
Every model can do that now. It is not impressive anymore. It is just a very polite brainstorm with turbo.
The interesting question was different:
Could I use AI to sort, criticize, cut away and arrive at an idea that could actually be tested?
The result was a prototype for Vært: a discreet Danish concierge for private dinners, hosting, culture, summer-house life and trusted services.
Not “ChatGPT for rich people”.
Thankfully.
The Problem Was Not A Lack Of Ideas
When you ask AI for app ideas, you quickly get a buffet.
And as with most buffets, the problem is not that there is too little food. The problem is that you end up with a little of everything and afterwards you are not quite sure why you took the pasta salad.
AI suggested all sorts of things:
- local AI assistant
- personal productivity coach
- app for young professionals
- social planning app
- summer-house app
- dining club
- private network
- digital concierge
- something with events
- something with premium access
- something with AI, because apparently everything needs AI in the name now
Some of it was good.
Much of it lived on the middle shelf of product development: not bad enough to discard immediately, but not sharp enough to build on either.
That is a dangerous zone.
Bad ideas die quickly. Mediocre ideas can live for a long time, especially if they have a nice name and a landing page.
The first lesson was simple:
AI is not most valuable as an idea machine. AI is most valuable as a filter.
The important question was not:
Can we invent more ideas?
Of course we could. A model can do that faster than a consultant can open a Miro template.
The important question was:
Which idea has a clear audience, a real problem, willingness to pay and a first version that can be tested without building an entire IT empire?
That is where the process became interesting.
The Broad AI App Died First
Many AI products start with the same sentence:
People need a personal assistant.
That is not wrong.
It is just too broad.
If everyone is the target group, nobody is the target group.
So I started pressing the ideas:
- Who has a problem that actually costs money?
- Who pays for quality and time, not just features?
- Where can Denmark be an advantage instead of a limitation?
- What can start manually?
- What becomes better with network, trust and local context?
- What should not be another chatbot subscription?
The last question matters.
We already have enough products where the concept is basically:
ChatGPT, but with our logo.
That can be an interface.
It is rarely a strong business by itself.
Slowly, one idea stood out:
A private Danish concierge.
Not a luxury app with gold UI, VIP language and desperate premium energy. More a calm service for people who can pay for time, discretion, taste and coordination.
The working title became:
Vært
Short. Danish. Calm.
Not Silicon Valley. Not Dubai. More Copenhagen, North Zealand and a summer house with the flowers somewhat under control.
The Better The Idea Got, The Less It Was About AI
The most interesting part of the process was that the idea became less and less “AI” as it became better.
At first, the easy pitch would have been:
An AI concierge for wealthy Danes.
That sounds like something from a pitch deck with gradients, stock photos and the word “seamless” too many times.
But the more I worked with it, the clearer it became:
The product is not AI.
The product is trust.
The product is access.
The product is saving time.
The product is being able to write:
We have six guests on Friday. It should be good, but not stiff. Find a solution.
And then receive an answer that actually fits.
Not a generic restaurant list.
Not a hallucinated wine menu.
Not a chatbot saying “that sounds like a wonderful evening!” before suggesting something that feels like Tripadvisor with confidence.
A real solution.
That also changed the role of AI.
AI does not need to be the front page.
AI can be the engine room.
It can help with:
- structuring requests
- summarizing preferences
- writing concierge briefs
- suggesting vendors
- drafting polite replies
- remembering what a member dislikes
- creating event checklists
- helping with research and partner outreach
But the experience itself should feel human, local and discreet.
That is a product lesson I think many teams will learn:
The more premium the experience, the less the AI should shout “look at me”.
Iteration Is The Real AI Work
There is a strange misunderstanding about AI: that the value is in the first answer.
It almost never is.
The first answer is raw material.
The good work comes in the loop afterwards.
I could keep challenging the model:
- “This is too generic.”
- “Find a narrower niche.”
- “Who actually pays for this?”
- “What is the first wedge?”
- “Why would it work in Denmark?”
- “What should we not build?”
- “Make it more exclusive, but not cringe.”
- “It should feel Danish, not American.”
- “Make a prototype without a backend.”
- “Cut anything that smells like a LinkedIn case.”
That is where AI becomes practical.
Not because it has perfect judgment. It does not.
But because it helps keep momentum in a thought process.
Many ideas normally die in the gray space between:
That could be fun.
and
What is version zero?
AI helps bridge that space.
But only if you stay critical.
Otherwise you just get a very well-written bad idea.
Prototype Before Platform
As a developer, you quickly want to build.
It is almost a reflex.
You see an idea and think:
- login
- database
- admin panel
- request system
- partner portal
- payments
- notifications
- maybe a little Kubernetes, because character flaws exist
But for Vært, that did not make sense yet.
The right first version was not a finished SaaS platform.
The right first version was a prototype that could explain the feeling:
- What is Vært?
- Who is it for?
- How do you send a request?
- What types of requests does it handle?
- What does the member experience feel like?
- Would anyone pay for this?
So the first prototype became static.
No backend. No database. No login. No paid AI API.
Just a showcase with:
- private member dashboard
- concierge request flow
- curated access
- partner and service network
- member preferences
- founding membership
It is not that backend will never matter.
It is that backend should be built when you know what it needs to support.
Otherwise you are building a very polished system for an assumption.
The Manual MVP Is Not Cheating
For developers, one of the hardest things is accepting that the best MVP might not be software.
It hurts a little.
You want a repo. A pipeline. An environment. A README. Maybe a small badge showing that something is green.
But Vært should be tested manually first.
Version zero can be:
- a landing page
- an invite-only form
- an email address
- a simple request process
- a spreadsheet with members and preferences
- a handpicked partner list
- manual coordination
That sounds primitive.
It is not.
It is product discipline.
If nobody will send requests manually, building an app will not help.
If nobody will pay for access, optimizing onboarding will not help.
If the requests turn out to be about something entirely different than expected, it is better to learn that in a spreadsheet than after three weeks of frontend work and a discussion about which toast library is most elegant.
What AI Actually Helped With
AI did not “invent” Vært from nothing.
That would be too neat a story.
AI helped make the process faster and more brutal.
It helped me:
- Expand the field of possibilities.
- Compare ideas against each other.
- Find weaknesses in the most charming ideas.
- Sharpen the audience.
- Turn a vague need into concrete use cases.
- Formulate a prototype.
- Hold on to MVP discipline.
- Say no to features that were mostly developer dopamine.
The last one is underrated.
AI can very easily give you more things to build.
The important use is getting it to help you build less.
The Next Test Is Reality
Vært is still only a prototype.
That is important.
A prototype is not a company.
A landing page is not traction.
A good idea is not proof.
And an AI-generated analysis is not willingness to pay.
The next test is much more practical:
- Can I talk to 10 relevant people?
- Will anyone say “I would actually use this”?
- Can I get founding members or strong verbal yeses?
- Can I log real concierge requests?
- Can I find reliable partners?
- Can the idea survive contact with reality?
That is where the project either becomes interesting or becomes another nice note in Obsidian.
And that is fair.
Obsidian can handle more notes.
The market is less sentimental.
Conclusion
AI is not a magical founder.
AI does not automatically make an idea good.
AI can absolutely help you build a bad idea faster, prettier and with more persuasive language.
But used well, AI can be a strong product partner.
Not because it knows everything.
But because it makes it easier to think in loops:
idea -> critique -> research -> structure -> prototype -> test
For me, the process did not end with another chatbot.
It ended with a sharper thought:
Wealthy Danes do not need more software. They need better access, less friction and someone who can make things happen tastefully.
That may become an app.
It may become a service.
It may first be only an email address and a good list of people you can trust.
But it is a better start than a generic AI assistant with a subscription.
Sometimes that is the point of AI:
Not getting it to build more.
Getting it to help you cut away until the idea becomes clear enough to test.
Read also: