Kevin Rose is trying once more. The founder of the original Digg, a site that once defined viral news through user votes, has rebooted the brand yet again. This time the focus narrows to AI news. No community posts. No upvote battles. Instead, the site watches what 1,000 influential voices on X discuss. Then it ranks stories by how fast they gain traction.
The new version lives at di.gg for now. Rose previewed it on May 8 with a simple X post. "a little project i've been hacking on," he wrote. He added a link and two words: "bugs expected." That candor fits the current state. The site feels raw. Yet its approach reveals something about the state of news consumption in 2026.
From Reddit Clone to Signal Detector
Last year's effort ended in disappointment. Digg relaunched as a Reddit rival early in 2026. Bots flooded the platform within hours. Sophisticated AI agents gamed the voting system. Engagement metrics collapsed under the weight of spam. The company laid off staff in March. CEO Justin Mezzell cited the "brutal reality" of the current environment in a note that The Verge reported.
Rose returned to the project full time in April. He ditched the social features. The new Digg pulls real-time data from X. It tracks views, likes, comments, quotes, and bookmarks. Then it applies sentiment analysis, clustering, and signal detection. The result appears on the homepage as ranked stories with visible metrics. One story might show 4.7 million views. Another claims the top spot for rising discussion.
Four highlighted sections sit at the top. Most viewed. Rising discussion. Fastest climbing. And the nostalgic "In case you missed it." Below runs a full daily list. Each entry includes X-derived numbers and links to the voices amplifying it. On May 11, top items included OpenAI's new Deployment Company, Thinking Machines' real-time collaboration models, and Cognition AI's Devin reaching a $445 million revenue run rate. The site lists specific X accounts by rank, from high-influence posters down to number 1,000.
Rose has spoken about this direction before. In a 2025 CNET interview, he said he would "heavily lean into AI on this front — AI for vetting and AI for a bunch of different things." Back then the plan still included community elements and advice from Reddit co-founder Alexis Ohanian. Those social ambitions have been set aside. The current product bets everything on curation through observation of existing conversations.
And the data flows constantly. The site reports live ingestion from X. It maintains graphs of connections. Clusters form around related posts. Sentiment scores help separate positive momentum from controversy. This creates a different kind of front page. One less shaped by what users submit to Digg itself. More shaped by what experts already share elsewhere.
The shift makes sense given recent history. Digg's earlier beta struggled precisely because it invited direct participation. Spammers saw the site's lingering Google authority as valuable. They struck fast. The new model removes that invitation. It observes instead. No accounts to ban. No votes to manipulate. Just patterns in public X data.
Yet questions remain. Will users return to a standalone site for AI news when X itself delivers a firehose of the same information? The "For You" feed already surfaces popular AI content. Many professionals follow the top voices directly. Digg adds analysis. Charts. Rankings of influencers, companies, and politicians active in the space. It promises to cut through noise. But does it add enough value beyond what careful X list management already provides?
Sarah Perez at TechCrunch captured the tension. The site might appeal to "data nerds" who enjoy seeing charts of how a Sam Altman interaction sparks wider discussion. Everyday readers could find less reason to visit regularly. Especially since no conversation happens on Digg. Users click through to sources or X threads. The platform acts as a sophisticated pointer rather than a destination.
Still, the approach addresses a genuine problem. AI news moves at blinding speed. Papers drop. Companies announce acquisitions. Researchers share preliminary results. Threads explode then vanish. Traditional aggregators struggle to keep pace. RSS readers deliver everything without priority. News apps often lag or push notifications based on their own editorial choices.
Digg's system claims to identify what thoughtful participants actually notice. It ranks by acceleration, not absolute volume. A story gaining discussion among mid-tier voices might surface alongside one boosted by Altman. The 1,000-voice list forms the foundation. Watch them. Measure what rises. Surface it. Simple in concept. Complex in execution.
Rose knows the stakes. He watched Digg rise in the 2000s, sell to Betaworks, fade, then get revived multiple times. The 2025 purchase with Ohanian carried high hopes. They talked about building for humans in an AI-filled web. They planned tools to verify real users. Now the product itself relies heavily on machine analysis of human chatter.
That irony sits at the center. Digg once thrived on direct human votes. The crowd decided the front page. Today the crowd posts on X. Machines watch the crowd. Machines decide what to highlight. The humans remain essential. Their attention and commentary provide the signal. But the sorting happens elsewhere.
Expansion plans exist. AI serves as the test case. Success here could lead to other verticals. Yet the method depends on concentrated discussion on X. AI attracts exactly that. Tech leaders, researchers, and enthusiasts post there constantly. Other fields have migrated. Threads, private Discords, and closed communities capture much non-tech conversation. Whether the model translates remains unproven.
Publishers might welcome the traffic. Google AI Overviews and changing search algorithms have slashed referral clicks. A site that drives engaged readers to original reporting could help. Especially if Digg's rankings highlight substantive work over pure hype.
The current alpha shows promise in its transparency. Metrics appear openly. Users see exactly why a story ranks high. They can trace influence back to specific accounts. This data-first presentation differs from black-box feeds on major platforms. It invites scrutiny. And with the site still labeled buggy, Rose seems comfortable letting observers watch the experiment unfold.
Whether this version sticks depends on execution. The team must refine the algorithms. Expand beyond AI without losing focus. Prove that the signal they extract beats what users already get from X or specialized newsletters. Past attempts at Digg revival stumbled on execution and market fit. This iteration feels more aligned with the original aggregator spirit. But the information environment has changed dramatically.
Rose bet that thoughtful voices still matter. That someone needs to watch them systematically. That acceleration of attention reveals importance better than raw popularity. The coming weeks will test those assumptions. For an industry drowning in AI content, even partial success could point toward better ways to stay informed. Or it could become another footnote in Digg's long story of reinvention.
WebProNews is an iEntry Publication