Skeptic intelligence is important in the age of AI

In meeting rooms, strategic offsites, and investors' summits, conversations always turn to artificial intelligence. Do we need our work, recharge our growth and expose hidden risks that we didn't expect? In the hype, one truth emerges. In a world caught up in machine-generated insights, there is the human ability to question, investigate and test.

But despite billions of spending on leadership development, few executives can accurately define the meaning of thinking skeptical, let alone how to develop it. To understand why skeptical intelligence deserves a seat alongside IQ and emotional intelligence, we need to revisit how these previous concepts shaped our understanding of human abilities.

IQ age

For most of the 20th century, intelligence only made sense of one thing. IQ. It was a gold standard, a quantifiable metric in which students were sorted, employees were promoted, and national rankings were compared.

The concept of general information was born in 1904 by Charles Spearman. He observed that individuals who performed well on one type of cognitive test tended to do well with other types. This statistical correlation suggests broad and fundamental mental competence. Alfred Vignette of France and later Lewis Turman of Stanford created IQ tests that could represent this ability numerically, leading to the IQ boom of the 20th century.

IQ has proven to be extremely good at predicting certain types of success. Academic performance, logical problem solving, and even long-term revenue. However, by the 1980s, cracks began to appear. Why did the average IQ flourish, and why did there have been top scorers in the real world of flounder?

The rise of emotional intelligence

The first serious challenge came from Howard Gardner, whose 1983 book framework introduced multiple theories of intelligence. Gardner argued that musical, spatial, kinesthetic, interpersonal, and intrapersonal intelligence are as realistic and valuable as linguistic or logical skills. This pluralistic view was controversial, but set the stage for a more focused alternative.

In 1990, two psychologists, Peter Saroby and John Mayer, proposed the concept of emotional intelligence. They defined it as the ability to effectively perceive, understand, manage and use emotions. In their view, emotions were not a distraction from rational thinking, but an important element of it.

But it was Daniel Golman who really set the global conversation on fire. His 1995 Bestselling Emotion Intelligence: Why more important than IQ is that self-awareness, self-regulation, empathy and social skills often outweigh leadership and raw cognitive horsepower in life. Goleman's work popularized the idea that high EQ can distinguish between great leaders and merely smart leaders.

Companies were enthusiastic about this concept. Emotional Intelligence Training has become the standard fare for GE, IBM, and Goldman Sachs leadership programs. The consulting company has developed all the practices regarding formula measurement and development. Yet, despite these advances, the dominant paradigm focused on how well we felt and connected.

Enter your skeptical intelligence

New concerns have emerged over the past few years. As machine learning systems become able to do amazing feats (cover legal briefs, diagnose illness, predict consumer churn), our natural tendency is to trust them. After all, algorithms look less biased than us, less emotional and more data-driven. However, recent famous obstacles – facial recognition systems that could not recognize skin skin faces, loan algorithms that punish women, and references hallucinating language models highlight AI as deeply flawed.

And these flaws are often subtle and buried in complex statistical models that even their creators struggle to fully interpret. result? The need for a new kind of human intelligence: the ability to critically interrogate the output of sophisticated systems. This is where skeptical intelligence appears.

Skeptic intelligence is not the same as mere anti-opposition or reflexive doubt. This is a disciplined approach to asking questions, combining curiosity, critical thinking, epistemological humility (knowing what you don't know), and a toolkit for assessing evidence. When IQ is about solving a well-defined problem, and EQ is about navigating social and emotional landscapes, skeptical intelligence is about resisting simple answers and investigating beneath the surface, especially when powerful technology tempts us to outsource our judgments.

Decades of research in critical thinking and cognitive psychology can be used to sketch its potential elements. Scholars such as Robert Ennis, Richard Paul, Rita McGrath, Eric Reis and Linda's Elders have long studied the implications of thinking critically. Their framework highlights the following capabilities:

• Clarification of concepts: Can you define what terms and recommendations actually mean? Many AI dashboards use ambiguous language (“potential customer engagement”) that lead to unfair interpretation.
• Looking for evidence: Do you naturally ask, “What data does this claim support?” “How was this data collected?”
• Identifying Assumptions: Are there any cultural, statistical, or organizational hidden facilities that need to surface?
• Consider alternatives: Ask “What other possible explanations or predictions?”
• Detecting cognitive bias: Are you wary of confirmation bias (preferring information supporting existing beliefs) or availability bias (overinformation of recent or vivid events)?
• Source evaluation: Not all experts and algorithms are equally reliable. Skeptic intelligence involves knowing how to scrutinize both human and machine authority.

In this sense, skeptical intelligence can be thought of as the nature of critical thinking that is strictly applied to modern data and AI landscapes.

Why is skeptical intelligence necessary now?

Paradoxically, the more AI you get, the more appealing it is to leave a skeptical faculty. Machine learning models often produce outputs with reliability scores and impressive graphs. A 2022 survey by Harvard Business School found that managers are very likely to accept flawed AI recommendations, even when discrepancies become apparent, if a visually compelling dashboard is presented.

This is not just a theoretical risk. Consider the 2020 incident in which the widely used recruitment algorithm in Fortune 500 companies found that it downgrades resumes from women, as training data contained historical biases in favor of male candidates. Or a set of fintech apps that misconceive minority borrowers as high risk based on opaque clustering techniques. These failures came not because executives were malicious or incompetent, but because they didn't have enough skeptical intelligence to interrogate the models.

Warren Buffett famously said, “It's good to learn from your mistakes. It's better to learn from other people's mistakes.” In the age of AI, it is best to be completely ahead of mistakes by fostering a healthy culture of skepticism.

This does not mean ignoring AI insights. Rather, it means creating a system that “trust validates.” Skeptic and intelligent leaders know how to ask data scientists pointless questions and challenge their assumptions without falling into endless analytical paralysis.

Skeptic Intelligence Practice

Imagine a CFO reviewing AI-driven forecasts that predict a 12% increase in demand for a new product line. Instead of simply praising or stacking rubber recommendations, the CFO trained with skeptical intelligence asks:

• Did the model weigh the largest?
• How similar were the training data conditions to today's market?
• What do you need to fake this prediction?
• How robust are these results for small changes in input?

Or imagine a Marketing VP using a generation AI tool to create campaign messages. Those with strong skeptical intelligence don't just check grammar. They cross-checked built-in stereotypical probes, tests for consistency of multiple prompts, and factual claims.

Skeptic intelligence also means knowing when to meet external experts, when to run pilot tests before a full-scale rollout, or when to keep humans in the loop for an appeal for ethical or reputable interest decisions.

Building skeptical intelligence in an organization

How can businesses today develop this new form of intelligence?

1. Training Beyond Compliance: While most companies offer general ethical or critical thinking modules, they rarely embed robust skeptical research into daily workflows. Scenario-based workshops (testing AI output under different conditions) can make this practical.
2. Employment for Epistemology Humility: Show curiosity in interviews by looking for candidates who can say “I don't know” and asking about company assumptions and models.
3. Constructive Opponents of Reward: Companies like Bridgewater Associates have long emphasized “thoughtful disagreements.” Encourage employees to question data-driven decisions. Especially when GroupThink sets it up.
4. Creating an AI Audit Team: Just as financial audit became the norm in the 20th century, algorithmic auditing could become an important feature of the 21st century. Teams dedicated to investigating how models are constructed and generalized can institutionalize skeptical intelligence.

Sceptical intelligence as a superpower

As historians look back at the early decades of the AI revolution, they may be amazed at how humans can easily be repeated by machines. Prosperous leaders will become people who speed up innovation, interrogation and scrutiny.

IQ and EQ remain fundamental. But skeptical intelligence – a disciplined, curious, humble ability that even the most clever systems question – proves to be the crown jewel of human abilities in the age of algorithms.