Diversity and inclusion problems in the Artificial Intelligence (AI) scene continue to be uncovered. And these are not to be taken lightly, especially because of their life-changing consequences.
For instance, an African-American man was falsely arrested due to a mistake made by a facial recognition system. Meanwhile, some video conferencing tools’ transcription features have a harder time transcribing virtual meetings of people whose native language is not English.
John Schauf, TaskUs Director for AI Operations, explained in his article published on Asia’s largest tech media platform e27 that “AI does not understand everything; it is only good at finding and recognizing learned patterns based on the training data given.”
“This overall lack of diversity in the industry results in a similar lack of awareness and knowledge about minorities’ issues,” Schauf said. “Having a diverse team introduces a wider range of perspectives and ideas, reducing the risk of bias and diversity issues from the get-go.”
He added that the more diverse machine learning teams and data labelers are, the more inclusive the technology will be.
Schauf’s e27 article is available in full here.