News for hackers. Stuff that matters.
Re:An ex-YouTube whistleblower on their recommendation engine by kodablah on Sunday February 10, 2019 @04:46PM (#19130257)
You can design for more than one thing, especially when complimentary. Why is there no room to speculate that they also designed for what people want? Why do I have to question the author's correctness to do so? Why mix all these words like this and tell other people what they can or can't speculate about?
Re:An ex-YouTube whistleblower on their recommendation engine by aznumeric on Sunday February 10, 2019 @06:23PM (#19130810)
Because the commenter did not add to the author's statement, but offered a different view on how YT's AI was designed. This particular comment section comments on the article, it is not a general discussion on YT's algorithm. Therefore if the author of the article, who is also one of the authors of the AI in question, writes a very clear statement on how the AI was designed, and the commenter then writes that it sounds like the AI is designed differently, then how can that mean anything else but questioning the correctness of the author's statement.