The Limits of Artificial Intelligence
The Limits of Artificial Intelligence
Blog Article
At a lecture hall in Manila, tech entrepreneur and investment icon Joseph Plazo made a striking distinction on what machines can and cannot do for the economic frontier—and why this difference is increasingly crucial.
Tension and curiosity pulsed through the room. Students—some eagerly recording on their phones, others streaming the moment live—waited for a man revered for blending code with contrarianism.
“AI will make trades for you,” he said with gravity. “But understanding the why—that’s still on you.”
Over the next lecture, he swept across global tech frontiers, balancing data science with real-world decision making. His central claim: Machines are powerful, but not wise.
---
Bright Minds Confront the Machine’s Limits
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. “Plazo’s words were uncomfortable—but essential.”
---
Why AI Still Doesn’t Get It
Plazo’s core thesis was both simple and unsettling: code can’t read between the lines.
“AI doesn’t panic—but it doesn’t anticipate,” 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
He didn’t bash the machines—he put them in their place.
“AI is the vehicle—but you decide the direction,” he said. It sees—but doesn’t think.
Students pressed him on sentiment tracking, to which Plazo acknowledged: “Sure, it can flag Reddit anomalies—but it can’t feel a market’s pulse.”
---
A Mental Shift Among Asia’s Finest
The talk sparked introspection.
“I believed in the supremacy website of code,” 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 understands not just volatility, but motive.
“Ethics can’t be outsourced to software,” he reminded. “Judgment remains human territory.”
---
An Ending That Sparked a Beginning
As Plazo exited the stage, students applauded. But more importantly, they stayed behind.
“I came for machine learning,” said a PhD candidate. “Instead, I got something more powerful—perspective.”
Perhaps, in drawing boundaries for AI, we expand our own.