top of page
  • EWC Community

Study Introduces the Possibility of Racist and Sexist Robots



By: Anna Luo


Last month, a study conducted by researchers around the world revealed that robots programmed with a popular artificial intelligence algorithm had sexist and racist biases.


Though this doesn’t impact modern day robots which are mainly used for manual labor, these biases could easily become problematic in the coming decade. Zac Stewart Rogers, a

supply chain management professor from Colorado State University, explains why. “With coding, a lot of times you just build the new software on top of the old software. So, when you get to the point where robots are doing more … and they’re built on top of flawed roots, you could certainly see us running into problems.”


So far, researchers have already begun seeing some of the aforementioned issues. Scientists used CLIP, a model that was released by OpenAI last year, to train robots in the study. The model is based off billions of online images and captions and is a much cheaper option that prevents robotics companies from having to create their own software.


In the study, when researchers asked robots to identify “criminals,” Black men were chosen almost 10 percent more than white men. However according to the scientists, the lack of sufficient background information should’ve prompted the robots to not respond at all. Similarly, women were less likely to be identified as a “doctor” than men and Black women were more likely to be identified as “homemakers” than White men.


Though there were limitations in the facial imagine data set the team used, and therefore could not account for nonbinary people, the study sends a clear message.


Due to the pandemic and subsequent labor shortage, companies are starting to rely more on automated workers. According to the Material Handling Institute, an industry trade group,“In the next five years, the use of robots in warehouses are likely to increase by 50 percent or more.”


Despite all this, Abeba Birhane, a senior fellow at the Mozilla Foundation, notes that it’s nearly impossible to develop bias-free robots. “This might seem radical,” she said. “But that doesn’t mean we can’t dream.”



Source:

- https://www.washingtonpost.com/technology/2022/07/16/racist-robots-ai/

5 views0 comments
bottom of page