Twitter partners with Mila AI research institute

Twitter wants to improve your experience on the service with the help of machine learning. The company has partnered with Montreal-based artificial intelligence research institute Mila to achieve that goal. The partnership's focus is Twitter's Cortex data and machine learning group. The three-year collaboration includes year-round open research engagements between the organizations, with Twitter researchers getting "elevated" access to Mila facilities and other internal resources. BetaKit reports there is a financial component to the deal to support Mila's projects, but the company hasn't disclosed the amount.

Twitter reverses decision to introduce tabbed timeline

Embed from Getty Images

People weren't happy with Twitter experimenting with an algorithmically-generated feed. The company planned to introduce a new tabbed interface that switches between the algorithmically-generated Home feed and the reverse-chronological Latest feed. Twitter rolled out this feature as a default for iOS users first, with plans to bring it to Android and the web after. But those who got the feature complained about how it defaulted to the Home feed every time they opened the app. 

Twitter said, "We heard you." And it decided to roll back the feature as the company "explore(s) other options." As Engadget pointed out, it seems many people online don't take too kindly to tech companies messing with the chronological feed. It somehow makes even more sense on Twitter, given it's a platform where many users get first-hand accounts and information on breaking news. So, an algorithm-based feed would make it challenging to access real-time information or reports.

Twitter might soon let you un-tag yourself in threads

Source: Jane Manchun Wong

Don't want to get notified in a Twitter thread anymore? You might be able to leave that conversation in the future. App researcher Jane Manchun Wong spotted a feature that allows you to leave conversation threads you don't want to be part of anymore. The "Leave this conversation" feature will let you un-tag yourself without letting the other participants know you aren't in the conversation anymore. They might still notice, given that they can't type in your handle. It'll just be displayed as plain text after you press the button. No one else will be able to tag you again in the same thread. That doesn't mean you won't be able to see the conversation. That's still something you can do. It's unclear if this feature will get a wide release. But if it does, is it something you find helpful?

Source | Via

Twitter expands Safety Mode beta test in Canada

Source: Twitter

After testing out its anti-abuse tool with a small group in September, Twitter has started rolling out the beta version of Safety Mode for more users across its English-speaking markets, including Canada. Extending testing means the company can get more insights into how the feature works and make necessary adjustments. Aside from expanding the rollout, Safety Mode can now prompt users when the app thinks you might need it enabled.

As TechCrunch wrote, Safety Mode is a defensive tool instead of one that proactively nudges conversations in the right direction. The idea behind the feature is to give users under attack a way to put up a defensive system without manually blocking each account harassing them. That can be difficult when a tweet goes viral, exposing the poster to potentially elevated levels of abuse. It doesn't just cover celebrities or public figures, but it could be members of marginalized communities, female journalists, and sometimes even ordinary folk.

Click to read more ...