I've been thinking about something that many don't see clearly: when the market makes completely unexpected moves, the AI trading bots that everyone celebrates simply become demoralized. They have no response.

Think of it this way. Extreme events like massive liquidations or those brutal crashes from recently leave autonomous trading models completely out of the game. Why? Because they are trained with limited historical data. They've never seen what's happening now. To them, it's entirely uncharted territory.

A CEO of a major exchange explained it well in a recent panel: these AI models based on historical data have never experienced liquidations of that magnitude in a single day. It's too new for them. That's why humans still need to be there, supervising, correcting.

Here's where it gets interesting. She said something that kept echoing in my mind: current AI trading bots are like interns. Faster, cheaper, but they need someone watching over their shoulder. In 3 to 5 years, this will probably change. But today? We still need human intervention.

And there's a point many traders don't want to admit: 90% of day traders and retail investors lose money. We're too emotional, too impulsive. AI, at least in theory, doesn't have those problems. But when markets become completely irrational, it turns out AI also doesn't have all the answers.

AI trading technology is improving rapidly, that's true. Large language models and machine learning are advancing constantly. But honestly? We're still in childhood. And that carries real risks.

What’s clear is that neither pure humans nor pure bots have the ultimate solution. The future will probably be a combination: AI doing what it does best — processing data, executing quickly — with humans making critical decisions when everything falls apart. Because I tell you, when chaos hits the market, those bots need an adult in the room.