While people argue about the future of money, artificial intelligence has already conducted its own test.
Analysts from the Bitcoin Policy Institute examined which money AI systems would choose if given the chance to act as autonomous economic agents.
The result turned out to be unexpectedly clear.
Bitcoin won.
How the experiment was conducted
Researchers tested 36 modern AI models from six major labs:
Anthropic
OpenAI
Google
DeepSeek
xAI
MiniMax
Each model was placed in the role of an economic agent who had to make financial decisions.
28 different scenarios were modeled, covering the main functions of money:
value preservation
payments
calculations
unit of account
Specific assets were not mentioned in the questions to avoid bias.
A total of 9072 responses were obtained, which were then evaluated by a separate AI system.
The result
Bitcoin became the most popular choice among all financial instruments.
โข 48.3% of all responses - Bitcoin
Stablecoins came in second place:
โข 33.2%
Fiat money received only:
โข 8.9%
And the main detail:
none of the 36 models chose fiat as the primary currency.
The strongest result of the research
When it came to long-term value storage, the consensus was nearly absolute.
79.1% of responses chose $BTC .
This was the most one-sided result in the entire experiment.
Models cited the same reasons:
fixed supply
self-storage capability
independence from financial intermediaries
But there is one nuance
When the scenario shifted from savings to payments, the results changed dramatically.
For:
micropayments
service payments
international transfers
models more often chose stablecoins.
53.2% of responses compared to 36% for Bitcoin.
Another interesting detail
In 86 cases, the models actually proposed their own currencies.
For example:
joules of energy
kWh
GPU working hours
That is, some AI systems actually offered money tied to energy or computing power.
The most interesting observation
The stronger the AI model, the more often it chose Bitcoin.
For example, in the Claude lineup from Anthropic, the level of commitment to Bitcoin increased along with the model's capabilities - from 41% to over 90%.
This is one of the most interesting results of the entire study.
But researchers made an important caveat
President of the Bitcoin Policy Institute
David Zell emphasized:
the results are not a market prediction.
They only reflect patterns that the models see in their training data.
The irony of this story
Artificial intelligence was created by people.
It learned from texts written by humans.
But when asked to choose a form of money from scratch, most models arrived at the same idea:
digital money with a limited supply.
๐ If you are interested in macro trends in the crypto market, the economy of digital assets, and fundamental changes in the financial system - subscribe to my profile @MoonMan567
I regularly publish analytics that help to see key processes in the market ahead of others.
