Gregory's ServerNeedless to say, using automated bias machines should be outlawed. We don't want it from people, and we certainly don't want to from more people, at a faster rate, mediated by machines, with no warnings and no recourse. Ugh. Why is this even something that needs to be said.
#ButlerianJihad
|
Top-level
 4 comments
I have smashed together two emojis to perfectly illustrate my reactions to the achievements of the artificial intelligence community! (Please note that I majored in biology and "new" artificial intelligence… in the last millennium.) |
For those that didn't read through, I find this the most revealing: "Retorio’s AI was trained using videos of more than 12,000 people of different ages, gender and ethnic backgrounds, according to the company. An additional 2,500 people rated how they perceived them in terms of the personality dimensions based on the Big Five model. According to the the start-up the AI‘s assessments have an accuracy of 90 percent compared to those of a group of human observers."
So what you now have is an automated processes to reproduce the biases of 2500 people. You can get your biased judgements so much quicker now! It's an efficient prejudice machine.
For those that didn't read through, I find this the most revealing: "Retorio’s AI was trained using videos of more than 12,000 people of different ages, gender and ethnic backgrounds, according to the company. An additional 2,500 people rated how they perceived them in terms of the personality dimensions based on the Big Five model. According to the the start-up the AI‘s assessments have an accuracy of 90 percent compared to those of a group of human observers."
So what you now have is an automated...