Twitter steps up initiative to stop the damage caused by its artificial intelligence
Twitter goes would have take a closer look at the “unintentional harm” caused by its algorithms. The recently announced initiative, dubbed the Responsible Machine Learning Initiative, has one primary focus. Namely, to take a close look at those algorithms that guide AI decision making across the service. More specifically, with regard to “racial subgroups” and gender.
What is the Responsible Machine Learning Initiative and how will it stop the AI bias in Twitter algorithms?
Now, Twitter is no stranger to prejudice. In fact, no other company works with an AI algorithm, but that’s what this Twitter initiative hopes to answer. Society has already admitted its racial prejudices, for example, in its image cropping algorithm. And he has tried to correct these types of errors where he finds them. But the problems are still there on that front, according to some reports. With algorithms favoring users with lighter skin.
And, in fact, this is one of the first problems listed by the company. Through its responsible machine learning initiative, Twitter hopes to take a closer look and analyze the cultivation system. This, according to Twitter, could also have some gender-related issues. And these will also be examined.
It will also conduct a “fairness assessment” with respect to its recommendations on the intake schedule, particularly with regard to racial subgroups.
Beyond race and gender, Twitter also hopes to analyze content recommendations based on political ideologies. These will be carried out in seven different countries.
How will you know when this initiative starts to have an impact?
Of course, no analysis or review would be remarkable if Twitter didn’t intend to do anything with the information. And, to that end, Twitter wants to apply everything it learns to improve the site and the app. For example, Twitter says the results may lead to the removal of algorithms. In effect, giving users more control over the images they choose to post on the site. Or it could translate into new construction and design standards AI algorithms in total.
But whatever path it takes, it also wants to educate its user base. In particular, not only on how algorithms work and what “informs” them, but also on their impact on what users actually see when they visit or use Twitter.
Additionally, the company plans to share any findings and solicit user feedback. That, the company says, will help it generate best practices for the industry and hold the company accountable.