When AI makes creation abundant, what is still scarce?
If building gets easier, then a lot more things get built.
If a lot more things get built, then a lot more things start to look substitutable.
And if substitutability rises, then the obvious question becomes:
what actually gets harder to replicate over time?
AI does not eliminate scarcity.
It changes where scarcity lives.
⸻
Scarcity didn’t disappear — it moved
A lot of AI discourse sounds like a story about abundance. And in one sense, it is. AI creates abundance in creation. In output. In variation. In implementation.
But abundance in one layer does not remove scarcity everywhere else.
It usually makes other constraints more important.
If building a product becomes easier, then the scarce things start to look more like:
•attention
•trust
•participation
•liquidity
•proprietary behavioral signal
•and, honestly, founder judgment
In other words: AI makes it easier to create software, but it does not make it easy to create the scarce resources that software depends on.
That shift is easy to miss.
It is also where a lot of durable advantage starts to come from.
Because if many people can build something, then the question is no longer just, who can create it?
The better question is:
who can capture something scarce through that product — and keep compounding it?
That is the lens I find most useful.
Not “what features can I ship?”
But:
•what scarce resource does this product touch?
•is it well-positioned to capture it?
•and does the system get stronger as it does?
That scarce resource might be:
•real participation in a network
•trust between counterparties
•actual market liquidity
•behavioral data that only emerges through repeated use
•user habits that deepen over time
•or outcome data that makes the next decision better
That is the real shift.
As creation becomes abundant, advantage moves toward capturing and compounding what remains scarce.
And that is exactly why network effects and feedback loops become more important in an AI world.
Not because they are buzzwords.
Because they are two of the clearest mechanisms for taking scarce things — participation, trust, signal — and turning them into accumulated advantage.
What AI makes cheaper:
writing code
making images
generating copy
answering questions
starting products
What AI makes more valuable:
taste
distribution
trust
proprietary data
judgment
@flutterflow @themoondesigner
The FF community doesn’t feel heard. Prove to us we’re wrong, and respond.
As a paying customer, it’s super disappointing to see what appears to be an abandonment of your community. Practically no new features are shipped anymore and what exists is barely maintained. Should we, as paying users, just cut our losses now and start building elsewhere? For example, after several months, FF still hasn’t updated to a Flutter version that allows testing in debug mode on physical device with iOS 26. Is that just never coming?!