Humans Already Screwing Up Robot Developments With Racism, Study Shows

Humans are pretty good at consistently doing the wrong thing.


Image via Getty/Caiaimage/Rana Dias


The abject shittiness of human beings' behavior, per a new study, may very well have an adverse effect on human-robot relationships and general interactions.

The Human Interface Technology Laboratory in New Zealand conducted a study dubbed "Robots and Racism," noting in the summary that the majority of robots currently being developed and sold "are either stylized with white material or have a metallic appearance." 

In the study, researchers utilized the shooter bias paradigm to see if participants were racializing robots, i.e. viewing robots as "black," "white," etc. As shown in the findings of the University of Canterbury-published study, racial bias against robots is an issue that needs addressing, with researcher Christoph Bartneck breaking down the importance of this finding in comments to CNN Thursday.

"It is amazing to see how people who had no prior interaction with robots show racial bias towards them," Bartneck said.

One experiment included in the study found that participants were quicker to shoot an armed black agent than an armed white agent, while also being quicker to refrain from shooting an unarmed white agent than an unarmed black agent, regardless of whether the agents were human or robot. These scenarios were virtually presented, with the robots shown on the screen possessing human-esque limbs and heads.

Bartneck told CNN, which also cited another study from earlier this year which showed that diversifying robots could reduce these biases, that society has "everything to win" by actively avoiding the development and sales of only robots that are presumed to be white.

Latest in Life