When scientists instructed specially programmed robots to look for “criminals” by scanning blocks with faces of humans on them, a surprising flaw was discovered. The block with a Black man’s face was regularly picked by the robots.
Researchers believe that these robots may be following in a troubling trend of robots engaging in sexism and racism. To answer questions, the robots filtered through billions of photographs and captions. Robots repeatedly selected blocks containing women and people of color in response to terms like “homemaker” and “janitor.”
Billions have been invested by businesses in the creation of robots to help replace people in jobs like stocking shelves, delivering packages, and even caring for hospital patients. Experts view the current climate for robots as something like to a gold rush, intensified by the pandemic as well as the ensuing labor crisis. But as the modern technology gets more sophisticated and commonplace, tech ethicists and experts are cautioning that widespread adoption of it could have unintended repercussions in the future.
A huge amount of data was used to train the robots in the study using an AI model CLIP. Unfortunately, the data sets are susceptible to human error, which could reinforce negative perceptions. There have long been worries about sexist and racially discriminatory AI algorithms. The nation-wide deployment of facial recognition technology by law enforcement leaves women and persons of color less easily recognized.
Andrew Hundt, a postdoctoral fellow from the Georgia Institute of Technology and lead researcher on the study, said this type of bias could have real world implications. imagine a scenario in which robots are requested to remove items from shelves. People are frequently depicted in literature, kid’s toys, and packaged food. He warned that if robots programmed with specific AI were used to select items, they would lean more heavily toward those featuring men or White people than others.
Robotics will play an important role in driving the automation industry’s $60 billion market price by the end of the decade. The use of robots in warehouses is anticipated to rise by 50% or more within the next five years. Amazon committed $1 billion to an innovation fund in April, which is heavily funding robotics firms.
According to Colorado State University supply chain management professor Zac Stewart Rogers, “With coding, you frequently just construct the new software on top of the old software.” “So, you could easily see us running into problems when you get to the point where robots are doing more… and they’re built on top of defective roots,” the author said.
Link: https://www.nytimes.com/2021/03/15/technology/artificial-intelligence-google-bias.html?searchResultPosition=1
Researchers believe that these robots may be following in a troubling trend of robots engaging in sexism and racism. To answer questions, the robots filtered through billions of photographs and captions. Robots repeatedly selected blocks containing women and people of color in response to terms like “homemaker” and “janitor.”
Billions have been invested by businesses in the creation of robots to help replace people in jobs like stocking shelves, delivering packages, and even caring for hospital patients. Experts view the current climate for robots as something like to a gold rush, intensified by the pandemic as well as the ensuing labor crisis. But as the modern technology gets more sophisticated and commonplace, tech ethicists and experts are cautioning that widespread adoption of it could have unintended repercussions in the future.
A huge amount of data was used to train the robots in the study using an AI model CLIP. Unfortunately, the data sets are susceptible to human error, which could reinforce negative perceptions. There have long been worries about sexist and racially discriminatory AI algorithms. The nation-wide deployment of facial recognition technology by law enforcement leaves women and persons of color less easily recognized.
Andrew Hundt, a postdoctoral fellow from the Georgia Institute of Technology and lead researcher on the study, said this type of bias could have real world implications. imagine a scenario in which robots are requested to remove items from shelves. People are frequently depicted in literature, kid’s toys, and packaged food. He warned that if robots programmed with specific AI were used to select items, they would lean more heavily toward those featuring men or White people than others.
Robotics will play an important role in driving the automation industry’s $60 billion market price by the end of the decade. The use of robots in warehouses is anticipated to rise by 50% or more within the next five years. Amazon committed $1 billion to an innovation fund in April, which is heavily funding robotics firms.
According to Colorado State University supply chain management professor Zac Stewart Rogers, “With coding, you frequently just construct the new software on top of the old software.” “So, you could easily see us running into problems when you get to the point where robots are doing more… and they’re built on top of defective roots,” the author said.
Link: https://www.nytimes.com/2021/03/15/technology/artificial-intelligence-google-bias.html?searchResultPosition=1
