Neighborhood social network Nextdoor is introducing new features powered by generative AI, the company announced Tuesday. Notably, Nextdoor is launching a new “Assistant” feature that helps users of the app write better posts. The new Assistant feature, which is being tested with select users, is powered by OpenAI’s ChatGPT.
The Assistant feature is designed to help users write posts that are more likely to receive positive community engagement. When writing a new post, users will see a revised suggestion that they can review and edit before finalizing their post. In addition, the Assistant will suggest ways to reword potentially harmful comments to make the same point in a friendlier way.
“With the launch of ChatGPT in November, we saw an opportunity to train ChatGPT technology on top of our own local knowledge to create something that we believe will really help users craft a more engaging message,” said Nextdoor CEO Sarah. Friar in an interview to AapkaDost. “We’re now giving you an assistant that says, ‘Hey neighbor, would you like to reword or reword this message so that more neighbors respond to it.’ It can help you post and ultimately get your neighborhood involved.”
Nextdoor currently has a “Kindness Reminder” feature that notifies users of the platform’s Community Guidelines and gives users a chance to think about a potentially harmful post. With the new Assistant feature, Nextdoor helps users rewrite messages to be friendlier. Say you and a neighbor are discussing the addition of new speed bumps in your neighborhood and you write a heated response. The Assistant will help you reword or reword your post to make sure your message comes across in a balanced and constructive way.
In addition, Nextdoor says the new generative AI technologies enable the platform to deliver more personalized and relevant content to users by analyzing the topic and context of conversations. This allows Nextdoor to better tailor content to users and provide more targeted recommendations for local news, events and activities.
Friar says Nextdoor is prioritizing security with this launch, noting that users don’t need to use the new features. For example, you don’t have to accept Assistant’s suggestions, you can go ahead and post as you please while still adhering to the platform’s community guidelines.
Nextdoor has initially started testing the new features with 25% of its users and is making them more widely available starting today, but the company has no plans to make the features available to 100% of its user base just yet. Friar says Nextdoor plans to continue testing the features and making sure the technology is rolled out safely before it’s made available to everyone. The features will roll out to everyone in the United States over time and eventually to the other markets where Nextdoor is available.
“We will be watching to make sure there are no unforeseen consequences, and of course accountability,” Friar said. “We will continue to move at the right pace to make sure we do it the right way. With any new technology there is always a concern about what might happen, but at the same time we want to make AI available to the masses. I feel everyone has the right to understand how this technology can make their lives more efficient.”
Nextdoor plans to continue testing generative AI models and is building a dedicated generative AI team led by the company’s head of AI, Qi He, who joined Nextdoor last year after more than eight years at LinkedIn . The dedicated team will hire product leaders and engineers to advance the company’s use of AI on its platform.
Last year, Nextdoor revamped its app with new profiles and more community-building features. The changes came as the social network developed a reputation for racial profiling over the years, leading to Nextdoor releasing specific features to address it and the company’s former CEO stepping down. The refresh was part of Nextdoor’s larger vision and goal to create a welcoming neighborhood both online and offline.