Twitch Set To Use AI To Stop Harmful Messages In Chat

Twitch Set To Use AI To Stop Harmful Messages In Chat

As someone who has spent countless hours streaming and interacting with the vibrant Twitch community, I must say this latest move by Twitch to incorporate AI in their chat system is both intriguing and slightly nerve-wracking.


In their latest update on X, Twitch Support has introduced a novel chat feature. This feature uses AI to review messages before they are sent, warning users if the content could potentially be considered offensive or disrespectful.

We’re introducing a new feature, driven by artificial intelligence, which nudges users to take a moment before sending chat messages that could be harmful. The aim of this trial is to minimize harassment and foster more constructive conversations on Twitch instead.

— Twitch Support (@TwitchSupport) August 12, 2024

Twitch just announced this as an experiment they’ll be running over the next few weeks.

Chat Hates AI

The Twitch Support post got some pretty harsh feedback, however. Many users question whether the AI will be able to distinguish between jokes, irony, and harmful words.

I noticed the use of the term “Butthead” being flagged on Twitch, and found it puzzling since it seemed quite innocuous. However, when I inquired about it with their support team, they clarified that even seemingly harmless words can be considered offensive in certain contexts, and therefore should be avoided to maintain a respectful and welcoming community.

Folks, we used the message example because, unsurprisingly, we did not want actual terrible language tweeted.

Stop being buttheads.

Fair enough.

Moderators will continue to manage chat discussions, but AI can serve as an assistant. It’s possible that AI could indeed contribute to making Twitch a more pleasant environment for everyone.

Read More

2024-08-13 17:09