Information Is Bad for Knowledge
More information can increase confidence without improving accuracy. Taleb uses image-recognition and forecasting studies to illustrate how early interpretations can persist even as additional details arrive.
THE IDEA
Information Is Bad for Knowledge
You may wonder how learning, education, and experience affect epistemic arrogance. The answer depends on the profession. The appearance of busyness reinforces the perception of causality, linking results to one's role in them. This applies to CEOs who trumpet their leadership as connected to company results.
Let us discuss one main effect of information: impediment to knowledge.
Aristotle Onassis, a shipping magnate, did not work in the conventional sense. He had no desk or office; his main tool was a notebook. He socialized, pursued women, and summoned lawyers to nightclubs at 2 A.M. There may be a "fooled by randomness" effect, but empirical research shows that additional knowledge of the minutiae of daily business can be useless, even actually toxic.
Empirical Evidence
-
Fire Hydrant Experiment: Show two groups a blurry image of a fire hydrant. One group sees it in ten slow steps, the other in five fast steps. Both are stopped at an identical image. The group with fewer intermediate steps recognizes it faster. Taleb’s interpretation is that the intermediate images encourage early hypotheses that are then hard to revise. Ideas are sticky; once you form a theory, you resist changing it. Two mechanisms: confirmation bias and belief perseverance.
-
Oskamp (1965): Clinical psychologists received increasingly detailed files about patients; their diagnostic accuracy did not improve, but their confidence grew.
-
Slovic: Bookmakers selected useful variables from past horse races; adding more variables did not improve prediction accuracy but increased confidence.
These accounts challenge the belief that "more is better": added detail can increase confidence without improving performance. The useful question is whether new information helps with the task.
When to use it
An investor who checks stock prices every minute may make poor trading decisions due to overreacting to noise; better to check less frequently.
A manager who micromanages with excessive data may miss the big picture and make biased decisions; focusing on key metrics can improve outcomes.
Checking whether additional detail improves a diagnostic or forecasting task, rather than treating greater confidence as evidence of better performance.
FROM THEORY TO EVERYDAY LIFE
In practice
Fire Hydrant Recognition Experiment
Subjects were shown a blurry image of a fire hydrant in either ten slow steps or five fast steps. Those who saw fewer intermediate steps recognized the hydrant faster, illustrating that additional information led to more hypotheses and worse performance.
Bookmakers' Overconfidence
Paul Slovic gave bookmakers sets of variables from past horse races. Adding more variables did not improve prediction accuracy but increased their confidence, illustrating a gap between confidence and accuracy.
A MOMENT TO REFLECT
Put it to work.
Think of a time when you had too much information and it clouded your judgment. How might you filter information to avoid overconfidence?
How does this concept apply to your consumption of news? Could reducing the frequency of updates improve your understanding?
Consider a decision you need to make. What is the minimum amount of information you need to make a good decision without being misled by noise?
References
- Clinical psychologists' diagnostic abilities with increasing information — Stuart Oskamp (1965)
- The influence of information on horse race bookmakers' predictions — Paul Slovic
CONTINUE THE CONVERSATION
Source texts
- The Black Swan The Impact of the Highly Improbable With a new section On Robustness and Fragility (Nassim Nicholas Taleb)