Technology will take down the Trolls | Unpublished
Hello!

Unpublished Opinions

mihaychuk's picture
Ottawa, Ontario
About the author

 I am interested in sustainability, civic engagement and democratic renewal. You might meet me at the soccer pitch, at the Farmers Market, or walking or biking on local trails.
I'm a high-tech veteran and have lived in Ottawa since 2000. Before moving to the private sector, I earned my Ph.D. in laser & semiconductor physics at the University of Toronto.

Like it

Technology will take down the Trolls

December 8, 2015

There has been a fair bit of coverage online and in a recent CBC The National panel on the issue of trolls in online comment forums. My comment presents a vision of how technology will evolve to fix the problem. This comment was specifically in response to a short article by Craig Hubley on LinkedIn (link to article - members only).

The CBC had to close its Comments section recently due to racist slurs against indigenous people, Muslims, and others. 

One alternative is to seek out niche forums or blogs where trolls do not ride coat-tails on large mainstream news media platforms. My favourite in the Ottawa area is http://UnpublishedOttawa.com. Trolls are less prevalent because they don't get the big audience.

The real solution to trolls will be more filtering, preferably automated text and troll-user recognition. Essentially, good anti-spam features for social media that make the tools work for ordinary, non-extremist, civil human users again.

It is a UX and human factors engineering problem, and AI will solve it. The platforms that don't solve it will mostly wither. A few creepy small platforms will always survive as smaller echo chambers for the lunatic fringe to talk to itself - but then, who ever really cared about the hateful fringe to begin with? The rest of us will "Keep Calm and Carry On", as the now-famous WWII poster says. 

The industry knows the problem. The problem has become big enough to limit the size of the market for online forums. It is time for the smarter technology providers to implement real solutions. The real solutions will need to be more ambitious than the older and simpler ways of human moderators or authenticating identities of users. The trick is to get to a system that is intelligent enough to enforce civility but not so intrusive that it enforces conformity, and banality or simply turns off users.

You can bet that Facebook and others are working feverishly on these features at this very moment.

*UX = tech-speak for "User Experience"
*AI = tech-speak for "Artificial Intelligence"