If AI Agents can truly “do things” for us in the future, the first things they may need to solve aren’t intelligence, but “identity” and “the ability to act.”
For the most part, the way we use AI today is still:
Ask a question → AI thinks → AI gives an answer → A human takes action.
But the real Agent Economy could look completely different:
You just need to tell an Agent, “Find me a hotel that fits my budget.”
Then the Agent needs to search, compare, communicate, and even complete the payment on its own.
Here’s the question:
How can an AI Agent prove “I am who I say I am”?
How can it obtain payment permissions?
How much can it spend?
What happens if it makes a mistake?
If there are thousands of Agents in the future, how will they conduct trusted transactions with one another?
That’s also why I’ve been paying attention to Kite lately.
As I understand it, Kite’s value isn’t just “giving AI a wallet.” It’s an attempt to solve the foundational problems behind the Agent Economy—giving Agents verifiable identities and enabling them to make payments and participate in economic activities within clearly defined permissions.
What’s truly interesting about this is:
As AI evolves from a “tool that answers questions” into an “Agent that can autonomously carry out tasks,” the dynamics of competition in AI may also change.
In the past, we compared:
Whose model is smarter?
In the future, we may also compare:
Whose Agent is more reliable?
Who can obtain permissions safely?
Who can transact with other Agents, services, and platforms?
If this trend takes hold, the Agent Economy may not simply be “AI + Crypto,” but an infrastructural shift in how machines participate in economic activity.
Of course, many questions remain unanswered:
How should permissions be controlled?
How should responsibility be determined?
How should fraud be handled?
How can Agents build trust with one another?
I think these questions are actually more worth discussing than “Will AI replace humans?”
Would you be willing to give your AI Agent some control over your payments?
#KiteBBW $KITE
#KiteBBW
For the most part, the way we use AI today is still:
Ask a question → AI thinks → AI gives an answer → A human takes action.
But the real Agent Economy could look completely different:
You just need to tell an Agent, “Find me a hotel that fits my budget.”
Then the Agent needs to search, compare, communicate, and even complete the payment on its own.
Here’s the question:
How can an AI Agent prove “I am who I say I am”?
How can it obtain payment permissions?
How much can it spend?
What happens if it makes a mistake?
If there are thousands of Agents in the future, how will they conduct trusted transactions with one another?
That’s also why I’ve been paying attention to Kite lately.
As I understand it, Kite’s value isn’t just “giving AI a wallet.” It’s an attempt to solve the foundational problems behind the Agent Economy—giving Agents verifiable identities and enabling them to make payments and participate in economic activities within clearly defined permissions.
What’s truly interesting about this is:
As AI evolves from a “tool that answers questions” into an “Agent that can autonomously carry out tasks,” the dynamics of competition in AI may also change.
In the past, we compared:
Whose model is smarter?
In the future, we may also compare:
Whose Agent is more reliable?
Who can obtain permissions safely?
Who can transact with other Agents, services, and platforms?
If this trend takes hold, the Agent Economy may not simply be “AI + Crypto,” but an infrastructural shift in how machines participate in economic activity.
Of course, many questions remain unanswered:
How should permissions be controlled?
How should responsibility be determined?
How should fraud be handled?
How can Agents build trust with one another?
I think these questions are actually more worth discussing than “Will AI replace humans?”
Would you be willing to give your AI Agent some control over your payments?
#KiteBBW $KITE
#KiteBBW