AI-powered news app Particle aims to reshape media dynamics

A startup called Particle, founded by former Twitter engineers, is aiming to transform the media industry’s relationship with artificial intelligence (AI) with a host of new reading experiences for users.

The AI-driven newsreader platform has already secured $4.4 million in seed funding from Kindred Ventures, Adverb Ventures and angel investors. The startup was conceived by Sara Beykpour, previously a Senior Director of Product Management at Twitter, alongside co-founder Marcel Molina, a former senior engineer at Twitter and Tesla.

Particle’s mission is not only to assist readers in understanding the news but also to support content publishers. Traditionally, AI in the media has had a contentious reputation, owing to its function in replacing reporters and cannibalising journalists’ work. Particle seeks to refashion this dynamic by ensuring that AI technology operates in harmony with the media industry’s business models.

From a consumer’s perspective, Particle offers a novel approach to news consumption. It is designed to enhance readers’ understanding of current events through various AI-powered features. More than just providing bullet-point summaries, Particle allows users to interact with news stories in diverse formats, including simplified summaries or audio narratives.

On the publisher’s side, Particle differentiates itself by seeking to redirect traffic back to news sites, rather than drawing from them without reciprocation. The app does this by prominently showcasing and linking to original sources directly within its AI-generated summaries. Notably, it has begun partnering with media outlets such as Reuters, AFP and Fortune. These alliances enable the hosting of select content within the app via the publishers’ application programming interfaces (APIs), offering partnered content enhanced visibility through distinct design elements.

Preliminary beta tests have shown that the app’s user interface influences readers to click through to the publishers’ content, reversing a trend where traffic might otherwise have been lost. With its public launch, Particle plans to deepen its collaboration with media partners, potentially offering users access to premium content, such as paywalled articles, under mutually beneficial terms.

Beykpour emphasises the company’s intent in fostering “deep partnerships and collaboration” within the media landscape. Among its strategies for traffic referral is a user-friendly design featuring large tap targets for easy navigation to publishers’ sites, as well as author profiles that guide users to additional related content.

Additionally, Particle incorporates built-in AI functionalities that introduce unique modes of news consumption. These modes include “Explain Like I’m 5,” providing layman-friendly explanations, and a focus on the “5 W’s” — who, what, why, when and where — to present straightforward factual summaries. It offers multilingual summaries and personalised audio options for on-the-go users.

In a bid to present news impartially, one of Particle’s standout features, Opposite Sides, attempts to dismantle echo chambers by showing diverse perspectives on the same topic. This tool illustrates stories on a spectrum, indicating the political slant of reporting from different outlets. By encapsulating both conservative and liberal leanings, it allows users to view summaries of these perspectives and form independent conclusions.

Perhaps the app’s most innovative feature is an AI chatbot that permits readers to pose questions about articles and receive fact-checked responses. Utilising OpenAI’s technology in tandem with Particle’s proprietary processing systems, the app can draw information from the web to supplement its responses with supporting evidence. The addition of human editors further ensures the accuracy and relevance of AI-generated content.

Particle utilises a variety of AI technologies, including OpenAI’s GPT-4o among others, along with traditional AI systems, to curate and enhance content. Beykpour asserts that the app’s systems significantly minimise potential inaccuracies in AI outputs, thereby maintaining high standards of information integrity.

Source: Noah Wire Services