Thread replies: 11
Thread images: 3
Anonymous
Diversity is strength, goodbye AI research
2016-06-26 20:31:04 Post No. 78842751
[Report]
Image search:
[Google]
Diversity is strength, goodbye AI research
Anonymous
2016-06-26 20:31:04
Post No. 78842751
[Report]
https://www.bloomberg.com/news/articles/2016-06-23/artificial-intelligence-has-a-sea-of-dudes-problem
Melinda Gates wife of Bill Gates, Margaret Mitchell a Microsoft researcher, and Margaret Burnett a professor at Oregon State University, and Lili Cheng who led the Microsoft team that developed Tay, argue there are too many men in AI and that this is a bad thing. Their concerns are lack of diversity, software biased towards rich white men, and different questions in research being asked. No evidence of any kind is provided. Arguments are based purely on guessing.
The author Jack Clark creates an argument for them by first by equating datasets and humans. Since dataset diversity is critical then human diversity must be critical. His evidence is Tay and blacks being classified as gorillas by accident.
His other arguments are simple predictions that if men train the AI systems they will be inherently flawed. These flaws will include not being able to understand ebonics and computers that act like men.
Jack Clark says job postings score high for masculinity in their speech. This is somehow bad. It ends with that if we train girls and women in AI they gain knowledge.
The implications are masculinity is toxic, diversity is strength, white men are rich, white men are biased, too many men are bad, and almost everything feminism espouses. Thanks to their politics, AI research will now proceed to be destroyed. This is what passes for journalism.