- Feb 3, 2015
- 90
Well in Zuck point of view, actually he doesnt care, and even if he has thousand of staff to review the content, i think AI manage billions of it, and if i the manager, i only ask my thousand staff to review if theres a complain about the page that has been removed but being requested to be enabled again or such if there any official complain.
Usually rarely the human staff will engage billions of page report everyday. Thus thats where the AI came to work, human reported the page until it reached the number of complains that match the certain limits where the removal of the page will start, then AI analyse the content of the page, and scan the possible cause, added to the database and when similar behavior post being posted, it automatically block the post and auto remove ads and auto remove the pages.
So in this kind of case, i will not blame Zuck, as one has played, test, experiment and also following others AI development around the world, AI can be effected by human behavior. Even Microsoft AI once want to start holocaust. Anyway it does not means it all bad, but it also does not means that AI are useless, AI are still under constant improvement from time to time so sometimes thats why Zuck cannot answer certain question unless he refer to the logs of AI action.
Even i as developer its sometimes hard to suddenly knew my AI goes wrong unless theres certain indicator that shows it is. Only then we will check the logs of its behaviour.
Usually rarely the human staff will engage billions of page report everyday. Thus thats where the AI came to work, human reported the page until it reached the number of complains that match the certain limits where the removal of the page will start, then AI analyse the content of the page, and scan the possible cause, added to the database and when similar behavior post being posted, it automatically block the post and auto remove ads and auto remove the pages.
So in this kind of case, i will not blame Zuck, as one has played, test, experiment and also following others AI development around the world, AI can be effected by human behavior. Even Microsoft AI once want to start holocaust. Anyway it does not means it all bad, but it also does not means that AI are useless, AI are still under constant improvement from time to time so sometimes thats why Zuck cannot answer certain question unless he refer to the logs of AI action.
Even i as developer its sometimes hard to suddenly knew my AI goes wrong unless theres certain indicator that shows it is. Only then we will check the logs of its behaviour.