Following its beta relaunch in January as a Reddit-like social news website and app, Digg shut down just two months later due to an influx of AI bots, leading to staff layoffs. But as of Tuesday, Digg is back. Again.

On the Digg homepage, Kevin Rose, who was part of the company's founding team in 2004, wrote that Digg is now an AI news aggregator at di.gg/ai. Rose signs his message with the CEO title, suggesting that former Digg CEO Justin Mezzell has moved on (though his LinkedIn hasn't yet been updated).

"Digg watches what 1,000 of the most thoughtful voices in AI are paying attention to, and ranks the stories they're pointing at by what's rising fastest," Rose said.

Called the Digg AI 1000, the ranked list of 1,000 people most involved in AI is built from the X social graph, which Rose describes as about 9 million follow relationships on X that map how the AI community connects. The Digg AI 1000 includes OpenAI CEO Sam Altman, Tesla and SpaceX CEO Elon Musk and Fei-Fei Li, a computer science professor at Stanford University, also known as the "godmother of AI." 

"Bugs expected. More topics soon," Rose wrote on X on Friday, sharing a preview of the new Digg site.

According to Rose's Digg homepage note, di.gg will be the home of its AI posts until the team is ready to officially launch on digg.com.

Digging through Digg

At the top of di.gg, there is a search bar, a dark mode button that turns the beige interface black, and two tabs: Stories and Rankings. Along with the 1,000 most influential voices in AI, Digg also tracks politicians and companies in the AI space under Rankings. 

Stories has a highlights section for the most-viewed, liked and commented-on X posts, along with an "in case you missed it" post. Below that, you'll find the day's top stories and those that are rising up the ranks, all of which have engagement metrics pulled from X, as Rose narrated in a screen recording posted to X on Thursday. 

Click on a post, and you'll see the X post driving traffic, the original post, a bar and percentage showing whether user sentiment is leaning positive or negative, how those in the Digg AI 1000 have interacted with it and "cluster engagement" line graphs for views, comments, bookmarks and reposts. Along with displaying engagement over time, these graphs show which Digg AI 1000 members interacted with the post and when. 

Posts that perform also appear to collect awards, such as "#1 liked" and "#1 engaged."

Digg's privacy policy

According to its privacy policy, in addition to X, Digg collects information -- specifically referencing public profiles -- from GitHub, a code project platform for developers. 

It also uses AI from services such as xAI, OpenAI, Anthropic and Google Gemini to "classify public accounts, summarize public posts and linked articles, describe public media, generate topic labels, score public content and power search." That would explain how Digg quantifies user sentiment into a percentage for its posts.

Along with public data from X and GitHub, the company might also collect visitor information, including IP addresses, and browser and device information. It also tracks the searches on its site.

Digg states that it doesn't sell personal information and only uses it to operate and improve Digg.