The moral bias behind your search results | Andreas Ekström

Share

Summary

Andreas Ekström explores the inherent bias in search engine results, particularly Google's, by distinguishing between searching for facts and searching for knowledge. He illustrates how search results can be manipulated and reveals Google's inconsistent intervention in such cases, highlighting the underlying human bias and the crucial need to integrate humanities with technology.

Highlights

The Illusion of Unbiased Search Results
00:00:13

When asked why they use Google, many students respond that it 'works' or that they're unaware of alternatives. Inevitably, some claim Google provides the 'best, unbiased search results.' This assertion, however, is philosophically almost impossible, as search engines often reflect underlying biases.

Fact vs. Knowledge: The Core Difference in Search
00:01:24

The speaker differentiates between searching for isolated facts, which Google excels at (e.g., 'capital of France'), and searching for knowledge. Knowledge involves bringing multiple facts together and valuing them differently based on individual perspectives, making it a far more complex and delicate endeavor that search engines are ill-equipped to handle objectively.

Case Study: Michelle Obama and Search Manipulation
00:02:55

In 2009, Michelle Obama was targeted by a racist campaign where manipulated images appeared in her Google image search results. This was achieved by purposefully labeling the images and files with her name. Google intervened manually to clean up these results and wrote code to prevent recurrence, deeming the results 'racist' and 'bad'.

Case Study: Anders Behring Breivik and Google's Inconsistency
00:05:14

Following the 2011 Norway attacks, a Swedish web developer, Nikke Lindqvist, organized a campaign to manipulate Anders Behring Breivik's image search results by associating his name with dog poop pictures. This campaign successfully pushed these images high in the search results. Notably, Google did not intervene in this case, unlike with Michelle Obama.

The Moral Bias of Algorithms
00:07:35

The speaker poses a critical question: why did Google intervene for Michelle Obama but not for Anders Behring Breivik? The answer points to a moral evaluation by Google – Michelle Obama is an 'honorable person,' and Breivik is 'despicable.' This reveals that behind every algorithm is a person with personal beliefs and biases. The speaker emphasizes the need for those in technology to identify and take responsibility for their biases, advocating for a stronger integration of humanities and technology to acknowledge that unbiased search results are a myth.

Recently Summarized Articles

Loading...