| |
Machine biases are patterns of mistakes in the logic of software such as algorithms or artificial intelligence. These are analogous to cognitive biases in people. Where human biases are well understood, the study of machine biases is in its infancy. Machines make mistakes in logic for several reasons:
CodingDesign errors, bugs or biases that are intentionally incorporated into software.Learning AlgorithmsMachine learning algorithms that cause an artificial intelligence to develop flawed models.Ensemble LearningEnsemble learning is a technique that combines results from multiple learning algorithms. It is similar to social processes and may contain biases that are analogous to social biases such as the abilene paradox.
Training DataArtificial intelligence that is trained using data that gives it the wrong impression about the world. For example, an AI that watches too many old Cowboy movies might assume that gun fights often break out during poker games.Outliers & Black SwansThe statistical models used by artificial intelligence may have difficultly processing outlier exceptions or small changes in certain parameters that result in major events. Humans may be better equipped to deal with surprising situations using skills such as creativity and synthesis.
Cognitive Biases
This is the complete list of articles we have written about cognitive biases.
If you enjoyed this page, please consider bookmarking Simplicable.
© 2010-2023 Simplicable. All Rights Reserved. Reproduction of materials found on this site, in any form, without explicit permission is prohibited.
View credits & copyrights or citation information for this page.
|