🧪 This platform is in early beta. Features may change and you might encounter bugs. We appreciate your patience!
apophenia
Apophenia is the tendency to perceive meaningful connections, patterns, or causal relationships in random or unrelated data. It encompasses pareidolia (seeing faces or figures in random visual patterns) and extends to finding spurious correlations in data, narratives in noise, and conspiracies in coincidence. It is a fundamental feature of human pattern recognition gone awry.
An investor notices that the stock market has risen on the last three Mondays in March and concludes there must be a 'Monday effect' in March, when in reality the pattern is purely coincidental and not statistically significant.
A sports fan notices their team has won every game this season on days when they wore their old college hoodie, and becomes convinced the hoodie is a lucky charm — rearranging their schedule to make sure they always wear it on game days.
A social media user sees three news stories about airplane incidents in one week and concludes that flying has suddenly become much more dangerous, not realizing that the algorithm is surfacing these stories based on engagement and that overall aviation safety statistics have not changed.
Binary (yes/no) questions an LLM must answer to identify this aspect:
Is a meaningful pattern, connection, or agent perceived in the data?
Type: binaryIs the data actually random, noisy, or unrelated?
Type: binaryWould a statistical test show the pattern is not significant?
Type: binaryApophenia is the tendency to perceive meaningful connections, patterns, or causal relationships in random or unrelated data. It encompasses pareidolia (seeing faces or figures in random visual patterns) and extends to finding spurious correlations in data, narratives in noise, and conspiracies in coincidence. It is a fundamental feature of human pattern recognition gone awry.
Humans evolved to detect patterns as a survival mechanism - seeing a predator that isn't there (false positive) is less costly than missing one that is (false negative). This hyperactive pattern detection persists even when dealing with genuinely random data.
Apply statistical tests to perceived patterns before acting on them. Remember that in any sufficiently large dataset, spurious patterns will emerge by chance, and ask whether a pattern would survive out-of-sample testing.
Apophenia drives conspiracy theories, superstitious behaviors in sports and gambling, and false discoveries in scientific research when researchers do not correct for multiple comparisons.
Use these tools to detect, analyze, or train this aspect.