Facebook is simulating users behavior using Artificial intelligence.
Facebook have recently developed a new program to help them identify and prevent harmful behaviors or scammers ,spammers or the buying and selling of drugs with the use of artificial intelligence powered bot by letting them loose on a parallel version of facebook.
These bot can then be studied and researchers can experiment on new ways to stop the act on facebook.
The research is being led by a facebook engineer called Mark Harman, Harman said WW was hugely flexible tool that could be used to limit a wide range of harmful behaviors on the site.
WW is a simulator that is based on facebook's real code base, they are a set of bot that is created and powered by facebook engineers on a parallel version of facebook to make themselves a target by creating a group of innocent bots to act as targets and a number of well trained bad bots to try to find them.by so doing, the engineers tried a numbers of ways to stop the bad bots like limiting the numbers of private messages and posts that the bot could send each minute, to see how this affected their behavior.
IN SUMARRY:The above system update on facebook is to reduce the way and tactics that scammers, spammers and hackers use to reach their targeted Fb users around the world.
Harman cites that the WW has it own limitations as it can't model the intent of users, the bots can not simulate complex behaviors, facebook says that the bot search, make friends request, leave comments,make posts and send messages but the actual conversation between users can not be simulated.
Harman says that the power of WW let facebook run thousands of simulations to check all sort of minor changes to the site without affecting users.
He also added that the main focus is training the bots to imitate things we know that happen on the platform.
Comments
Post a Comment