TECH THAT...

A ROUND-UP OF THE LAT­EST NEWS IN THE DIGITAL WORLD

Llanelli Star - - TECH NOW -

IN­STA­GRAM TESTS NEW FEA­TURES DE­SIGNED TO BEAT THE BUL­LIES

UN­TIL re­cently the founders of so­cial net­works have not cared much about what goes on, good or bad, on their plat­forms as long as the num­ber of users was ris­ing – but no more. They are re­ally start­ing to do some­thing about bul­ly­ing.

Ar­guably, this is nowhere more im­por­tant than on In­sta­gram. This week, its boss Adam Mosseri an­nounced two new fea­tures it is rolling out to try to make a dent in the bul­ly­ing cul­ture on the net­work – and most welcome they are too.

First is an AI mon­i­tor­ing of com­ments made, that will ask the poster to re­flect a lit­tle on what they’ve writ­ten and ask if they re­ally want to post it. In­sta­gram says in test­ing this has proved rea­son­ably ef­fec­tive – a lot of users, when try­ing to post a mes­sage or com­ment that of a type recog­nised as hate­ful by the AI, think twice and delete when asked the ques­tion.

In­sta­gram is also in­tro­duc­ing a sys­tem that al­lows a user to “shad­ow­ban” any­one they don’t like from com­ment­ing on their im­ages or send­ing them di­rect mes­sages.

Shad­ow­ban­ning a the bully means their com­ments are not made pub­lic un­til ap­proved. The bully can see them, but they don’t know them­selves that no one else can.

The bully’s di­rect mes­sages don’t show up to them as hav­ing been read, ei­ther, and they can’t see if the per­son they’re bul­ly­ing is on­line.

Es­sen­tially it means they think they’re just be­ing ig­nored, and no one else is able to pile on to val­i­date their be­hav­iour.

It sounds like a solid and so­phis­ti­cated way to deal with the prob­lem with­out pour­ing fuel on the fire. Let’s hope it shows re­sults.

Newspapers in English

Newspapers from UK

© PressReader. All rights reserved.