The AI Effect: Why We Always Move the Goalposts of Intelligence

Every time artificial intelligence achieves a milestone once thought to belong exclusively to human cognition, a strange psychological phenomenon occurs.

We don’t celebrate it as a triumph of machine intelligence. Instead, we shrug, diminish the achievement, and declare, “Well, that’s not really AI; that’s just math/search algorithms/pattern matching.”

This phenomenon—the sliding definition of what counts as artificial intelligence—is known as The AI Effect. It is perhaps the single most revealing quirk of human psychology in the digital age, exposing less about the limits of technology and a great deal more about human ego.

The Moving Horizon

The term was coined by computer science pioneer John McCarthy, who famously quipped: “As soon as it works, no one calls it AI anymore.”

History is littered with corpses of once-futuristic concepts that were evicted from the “AI” category the moment they became commercially viable or commonplace.

• Deep Blue (1997): When IBM’s chess-playing supercomputer defeated world champion Garry Kasparov, headlines declared it a watershed moment for machine intelligence. Yet, critics quickly pivoted: “It’s just brute-force calculation looking millions of moves ahead. It doesn’t actually understand chess.”
• Optical Character Recognition (OCR) & Spam Filters: Decades ago, teaching a computer to read handwriting or automatically sort junk mail from your inbox was cutting-edge machine learning. Today, it is treated as mundane infrastructure utilities. No one looks at a modern spam filter and thinks, “Ah, a marvel of synthetic cognition.”
• Generative AI & LLMs: Just a few years ago, the idea of a machine writing poetry, generating photorealistic art, or holding a fluent conversation sounded like sci-fi. Now that millions of people use these tools daily, the narrative has shifted to dismissal: “It’s just a glorified predictive text engine throwing probabilities together. It doesn’t actually ‘know’ what it’s saying.”

Why Do We Move the Goalposts?

The AI Effect is driven by a deep-seated human bias known as anthropocentrism—the tendency to place humanity at the center of the universe, particularly when it comes to our unique cognitive traits.

For centuries, humans defined themselves by what animals and machines could not do. We were the tool-makers, the chess players, the artists, the language users, and the logical reasoners. Every time a machine invades one of those territories, it feels like an existential deflation.

To protect our self-image, we engage in a psychological defense mechanism: redefinition.

• When a task requires immense effort for a human, we label it “intelligent.”
• When a machine performs that exact same task effortlessly via algorithms, we redefine the task as “mechanical” or “unintelligent.”

In short: If a computer can do it, it can’t require real intelligence.

The Double-Edged Sword of the AI Effect

While the AI Effect sounds like a harmless semantic debate, it has very real-world consequences.

1. The Hype-and-Despair Cycle

The AI Effect fuels the rollercoaster of public perception. When a breakthrough happens, we experience “AI Winter” panic or sci-fi utopian hype. But once the technology integrates into everyday life, public disappointment sets in because it doesn’t match the omnipotent, sentient artificial general intelligence (AGI) promised by Hollywood. This undervalues genuinely revolutionary tools that quietly transform medicine, logistics, and scientific research.

2. Undervaluing Human Labor

By dismissing AI outputs as “just math” or “stealing work without thinking,” we often overlook the human curation, prompt engineering, fine-tuning, and systemic alignment required to make these tools useful. Conversely, it can also lead society to underappreciate the complex cognitive layers embedded in jobs we traditionally thought of as “low-skill” (like content moderation or data labeling) once we realize machines still need human guidance to navigate nuance.

Redefining Intelligence

Ultimately, the AI Effect forces us to confront an uncomfortable question: What is human intelligence, really?

If chess, painting, coding, and translation can be reduced to patterns, probabilities, and optimization functions, what does that say about our own brains? Perhaps human intelligence is also a staggering, biological optimization engine—just running on a different kind of hardware.

Until we settle on a definitive answer, the goalposts will keep moving. The moment artificial general intelligence (AGI) finally arrives, society won’t marvel at the dawn of a synthetic mind. We’ll likely look at it, yawn, and say, “Of course it can do that—it’s just a computer.”