The Limits of Artificial Intelligence
The Limits of Artificial Intelligence
Blog Article
In a packed amphitheater at the University of the Philippines, Joseph Plazo laid down the gauntlet on what AI can and cannot achieve for the future of finance—and why understanding this may define who wins in tomorrow’s markets.
You could feel the electricity in the crowd. Students—some furiously taking notes, others capturing every word via livestream—waited for a man known not only as an AI visionary, but also a contrarian investor.
“Machines will execute trades flawlessly,” Plazo opened with authority. “But understanding the why—that’s still on you.”
Over the next hour, he swept across global tech frontiers, touching on everything from quantum computing to cognitive bias. His central claim: Machines are powerful, but not wise.
---
The Audience: Elite, Curious—and Disarmed
Before him sat students and faculty from leading institutions like Kyoto, NUS, and HKUST, united by a shared fascination with finance and AI.
Many expected a celebration of AI's dominance. What they received was a provocation.
“There’s too much blind trust in code,” said Prof. Maria Castillo, an Oxford visiting fellow. “We need this kind of discomfort in academia.”
---
The Machine’s Blindness: Plazo’s Case for Caution
Plazo’s core thesis was both simple and unsettling: AI does not grasp nuance.
“AI won’t flinch, but neither will it foresee,” he warned. “It finds trends, but not intentions.”
He cited examples like machine-driven funds failing to respond to COVID news, noting, “By the time the algorithms adjusted, the humans were already positioned.”
---
The Astronomer Analogy
Plazo didn’t argue against AI—but for boundaries.
“AI is the microscope—you choose what to zoom in on,” he said. It sees—but doesn’t think.
Students pressed him on AI in news and social chatter, to which Plazo acknowledged: “Yes, it can scan Twitter sentiment—but it can’t discern hesitation in a policymaker’s tone.”
---
The Ripple Effect on a Digital Generation
The more info talk sparked introspection.
“I thought AI could replace intuition,” said Lee Min-Seo, a quant-in-training from South Korea. “Turns out, insight can’t be uploaded.”
In a post-talk panel, faculty and entrepreneurs echoed the caution. “This generation is born with algorithmic reflexes—but instinct,” said Dr. Raymond Tan, “is not insight.”
---
What’s Next? AI That Thinks in Narratives
Plazo shared that his firm is building “co-intelligence”—AI that blends pattern recognition with real-world awareness.
“No machine can tell you who to trust,” he reminded. “Capital still requires conviction.”
---
Standing Ovation, Unfinished Conversations
As Plazo exited the stage, the crowd rose. But more importantly, they started debating.
“I came for machine learning,” said a PhD candidate. “But I left understanding myself better.”
In knowing what AI can’t do, we sharpen what we can.