I just came across the Square Top 100 Creator Awards, and it turns out the sole sponsor is @KITE AI 中文 . This seemed like a good opportunity to share my honest thoughts on the Agent economy.
This afternoon, I was glued to the 1-minute candlestick chart, making a quick futures trade and feeling completely on edge. Then a question suddenly came to mind: What would happen if, in the future, an AI Agent—not me—watched the charts and placed trades on my behalf?
Say I topped it up with 100 U to arbitrage, but there was a bug in the code, or it went haywire and spammed paid API calls, or repeatedly opened and stopped out of positions in a very short time, burning through my money like it was game tokens. Surely I can’t be expected to manually approve every single trade, right?
So, for AI to truly “get things done” on-chain, intelligence alone isn’t enough—it needs someone to keep its wallet in check. That’s what I find interesting about Kite.
I looked into it, and @KITE AI 中文 is essentially doing one thing: putting an “even tighter leash” on Agents. You can set rules, like a daily budget of 100 U, and if it goes over budget, it simply won’t have permission to spend more. It also supports high-frequency micropayments in stablecoins, so even a 0.01 U API call can be accounted for precisely. Add the identity layer, and when millions of Agents work together in the future, there’ll be a clear record of who did what.
Put simply, a large model is an Agent’s brain, while $KITE manages its wallet and keeps it within the rules. Without this underlying infrastructure, no matter how smart AI gets, we won’t dare entrust it with real money.
I’m using the Top 100 campaign as an excuse to pop in. I just hope Kite builds more real-world applications going forward, instead of making empty promises. Guys, how much of your portfolio would you trust AI to manage right now?
#KiteBBW $KITE
This afternoon, I was glued to the 1-minute candlestick chart, making a quick futures trade and feeling completely on edge. Then a question suddenly came to mind: What would happen if, in the future, an AI Agent—not me—watched the charts and placed trades on my behalf?
Say I topped it up with 100 U to arbitrage, but there was a bug in the code, or it went haywire and spammed paid API calls, or repeatedly opened and stopped out of positions in a very short time, burning through my money like it was game tokens. Surely I can’t be expected to manually approve every single trade, right?
So, for AI to truly “get things done” on-chain, intelligence alone isn’t enough—it needs someone to keep its wallet in check. That’s what I find interesting about Kite.
I looked into it, and @KITE AI 中文 is essentially doing one thing: putting an “even tighter leash” on Agents. You can set rules, like a daily budget of 100 U, and if it goes over budget, it simply won’t have permission to spend more. It also supports high-frequency micropayments in stablecoins, so even a 0.01 U API call can be accounted for precisely. Add the identity layer, and when millions of Agents work together in the future, there’ll be a clear record of who did what.
Put simply, a large model is an Agent’s brain, while $KITE manages its wallet and keeps it within the rules. Without this underlying infrastructure, no matter how smart AI gets, we won’t dare entrust it with real money.
I’m using the Top 100 campaign as an excuse to pop in. I just hope Kite builds more real-world applications going forward, instead of making empty promises. Guys, how much of your portfolio would you trust AI to manage right now?
#KiteBBW $KITE