News-gathering apps have been around for years. However, most of them have been edged out of the market by social media networks with algorithms that are better at understanding the nuances involved in how people consume news.
But now, tech companies are looking towards AI to help them deliver personalised news in a more valuable way. One such AI startup in the fray is Particle, co-founded in February last year by two former Twitter executives.
Particle’s AI-powered news-reader app of the same name recently grabbed headlines after a beta version was privately released for testing among a select group of iOS users.
What is new about Particle?
Particle’s driving philosophy is to not only provide AI-generated summaries of news but to do it in a way that helps users easily make sense of the news.
“Part of Particle’s thesis is how a story is being reported from all sides of the spectrum,” co-founder Sara Beykpour was quoted as saying by TechCrunch. “We wanted to build an experience for consumers that really helps them cut through the noise and that helps them understand more [and] what’s going on faster, and we want to do that in a way that is a sustainable, win-win situation for both the readers that are consuming and the publishers and the journalists that are producing the content,” she said.
Beykpour worked at X when it was known as Twitter. She was reportedly involved in projects such as Twitter Blue and Conversations during her stint at the social media giant between 2015 to 2021.
Marcel Molina, Particle’s other co-founder, is a former senior engineer who worked at Twitter as well as Tesla. While the duo initially wanted to launch an app with a social media focus, they later pivoted to news and information as a primary focus for the platform.
How does Particle work?
Particle reportedly uses a hybrid of AI models, including OpenAI’s GPT-4o, to primarily organise and summarise news articles.
To help users get a lot of information and varied perspectives on an issue, the app bundles news coverage into collections called ‘Stories’. Each Story can include multiple news articles as well as X posts. It also has a section with only quotes on a particular news topic.
The top of each Story carries an AI-generated summary of all the news articles below. It is unclear if this also includes hyperlinks to those news articles.
Notably, Particle allows users to change the style of the AI-generated summary. For instance, a user can request the app to give both sides of a news event. They can also ask Particle’s AI bot questions to help simplify topics such as the war in Gaza.
Users can have the headlines of news articles rewritten in a manner that is less confusing. Users can also follow specific news outlets or journalists on Particle.
The app reportedly selects news articles from a wide range of publishers in order to expose users to diverse viewpoints. It does not currently allow users to share news articles from their end.
Does Particle have permission to share publisher content on its app? The company recently inked deals with major news publishers such as Reuters, Time, and Fortune to display and summarise the news articles posted on these sites.
In June this year, Particle raised over $10.9 million after it closed a Series A funding round led by global VC firm Lightspeed Venture Partner. It also saw investments from global media house Axel Springer, which has organisations such as Business Insider and Politico under its belt.
Can Particle succeed where others have failed?
The app graveyard is littered with many news-aggregation platforms such as Circa and Discors.
Artifact, an AI-powered news app developed by the co-founders of Instagram, was on the verge of being taken down only to be acquired by Yahoo and integrated into its news app earlier this year.
News reader apps have broadly struggled to attract an audience by offering summarisation and aggregation features which dashed their opportunities to make any money out of it.
“We have built something that a core group of users love, but we have concluded that the market opportunity isn’t big enough to warrant continued investment in this way,” the makers of Artifact had written in a blog post. This was, of course, before their rescue boat arrived.
They had also pointed out how the content on Artifact had required “a fair amount of moderation and oversight” for which they did not have the staff.
However, Particle claims to be different wine in a different bottle. It plans on reducing the risk of AI hallucinations and inaccuracies through human editorial oversight as well as publisher deals.
Particle is also relying on its unique, multi-perspective approach to avoid having users encapsulated in their own algorithm bubbles. Whether it will be able to meet publisher needs and compete with AI-powered search engines remains to be seen.