The takeaways
- Google has launched a new social network named Keen for web and Android.
- The platform uses Google Search and human-centered machine learning AI to promote relevant content based on users’ personal interests.
- Users can create and share their curated resources with one another, which in turn educates both the user base and Keen’s AI on their preferences.
What happened?
Keen, a social networking platform by Google’s in-house incubator for experimental products, Area 120, quietly launched earlier this month, both online and on Android. Mooted as a rival to Pinterest, Keen aims to “curate, collaborate and expand everything” users are interested in by utilising Google’s machine learning expertise to promote increasingly relevant topics to Keen’s users.
Co-founder CJ Adams has promoted Keen as an alternative to “mindless” online browsing that focuses people’s attention on their hobbies and passions. In his post, he revealed that it was designed in collaboration with a Google team named People and AI Research (PAIR) that specializes in human-centered machine learning systems. Tech website Search Engine Journal also described the platform as “poised to become a new source of referral traffic, which makes it worthy of investigation”.
How does it work?
Users can create a ‘keen’ on any topic, curating their content and links on a pinboard similar to Pinterest’s popular interface. The main difference lies in the machine learning that enables greater personalization based on each keen’s content.
Powered by Google Search, Keen’s machine learning, much like Instagram’s algorithm, uses the data to show users more of what might also interest them. This includes following other keens and creating alerts for new content. Users are encouraged to share their keens with one another, thus promoting it as a referral, passion and knowledge-driven community. At present, Keen’s specific data policy is unknown. It is covered instead by Google’s standard privacy policy.
