Why AI Fails When the Market Panics
Why AI Fails When the Market Panics
Blog Article
AI trailblazer Joseph Plazo just reminded a room full of elite students something Wall Street won’t admit: AI may be powerful, but it’s not wise.
MANILA — He didn’t show up to sugarcoat things. He came to crack illusions.
On a sweltering Thursday morning at the Asian Institute of Management in Manila, Plazo stood before a sea of students from top Asian universities—AIM—ready for a sermon on AI’s glory in finance.
What they got instead? A jolt of truth.
“AI is like your smartest intern,” Plazo noted, “But you still don’t give the intern the keys to your vault.”
The room laughed. Then they stilled. Because he was dead serious.
### The Flaw in the Code: No Judgment
Let’s be clear—Plazo isn’t some technophobe clinging to the past. He builds trading AIs. His firm, Plazo Sullivan Roche Capital, creates some of the most accurate systems across global markets. He understands machine learning like few do.
But that’s why his warning felt urgent.
“The problem isn’t AI,” he told the room. “The problem is us. We keep believing it’ll save us from making hard decisions. That’s not its job.”
Plazo unpacked real-world case studies—moments when AI signaled winning trades… just ahead of a central bank pivot or an unexpected war. Moments no dataset could foresee.
### Asia’s Best Challenged Him—and Learned Something
A student from Kyoto asked if LLMs might someday gauge global sentiment.
Plazo answered without blinking.
“AI can spot a tweetstorm. But it won’t sense dread in a press conference. It won’t catch regret in a central banker’s sigh.”
The room exhaled. Message received.
Another asked, “Can AI ever understand conviction?”
Plazo raised an eyebrow.
“Conviction isn’t math. It’s instinct. It’s forged by failure and memory. You can’t download that.”
### Plazo’s Words = Financial Therapy
This wasn’t about flash trading or chatbots. It was about ethics.
Students admitted they saw AI as a cheat code—an escape hatch from risk, from thinking too hard. Plazo called it out.
“You can automate your trades. You can’t automate your values.”
That line slapped. Because everyone in that room—from the copyright cowboys to the quant whizzes—wanted alpha. But not at the cost of why they started.
### So What’s AI Good For?
Plazo didn’t trash AI. He credited its strengths:
- It filters noise.
- It backtests at scale.
- It spots technical setups better than any human.
But it can’t read sarcasm. It fails to sense when a politician is bluffing. And it doesn’t know if your retirement burns.
“If your AI bot makes a bad call,” Plazo asked, “do you still accept blame? Or do you blame the code?”
That was the mic drop.
### This Isn’t About AI—It’s About You
Plazo wasn’t preaching finance. He was preaching accountability. Use AI—but don’t worship it. Let it assist—not decide.
And yes—he still believes in the machines. He’s building tools more info that track geopolitics, misinformation, even psychological nuance.
But he left no doubt:
“No machine can tell you when *not* to act. That’s your job.”
### Final Thought: Maybe the Future Needs Less Code—And More Courage
As the crowd filed out—buzzing, challenged, changed—one phrase echoed down the halls:
“AI doesn’t know your values. So don’t let it make your decisions.”
In a world chasing speed, Plazo offered something rarer:
A mirror.
Because investing isn’t just about *winning*. It’s about knowing **why** you played.