A team led by a former Twitter engineer is reimagining how AI can be used to help people process news and information. Particle.news, which launched a private beta over the weekend, is a new startup that offers a personalized, “multi-perspective” news-reading experience, leveraging AI to not only summarize the news, but also to present it in an unbiased way. aims to summarize. The same goes for compensating or claiming authors and publishers.
Particle has yet to reveal its business model, but it arrives at a time of growing concern about the impact of AI on the rapidly shrinking news ecosystem. News summarized by AI could limit clicks to publishers’ websites, which means their ability to make money through advertising would also be reduced.
The startup was founded last year by Sarah Beikpour, former senior director of product management at Twitter. Sarah Beikpour worked on products such as Twitter Blue, Twitter Video, and Conversations, and spearheaded the experimental app twttr. She remained at Twitter from 2015 to 2021, rising through the ranks from software engineering to senior director of product management. Her co-founder is Marcel Molina, a former senior engineer at Twitter and Tesla.
As Beykpour explained last month, the premise behind Particle is to use AI to make it easier to keep up with the news.
“Sometimes it feels like all of our time is writing headlines. We also want to understand more, faster,” she writes in the startup's introduction to Threads. I am. “We are in the early stages of using AI to transform the way we interact with news.”
Particle provides news readers with a quick, bulleted summary of the article, including information pulled from a variety of sources. But when Beykpour announced the private beta, he said readers could use the summary to get an overview or dig deeper to “learn how the story has developed over time.” He said he could.
The venture-backed startup has raised a total of $4.4 million in seed funding from Kindred Ventures and Adverb Ventures, as well as a variety of angel investors including Twitter and Medium co-founder Ev Williams and Behance founder Scott Belsky did. This round he finished in April 2023.
Remarked Belsky on X“Particle has become an app I use every day. It consolidates many articles (and angles) on any news topic, brings out the important points as objectively as possible, and allows me to dig deeper from different aspects. “This is a great example of everyday AI in the age of abstraction that lies ahead,” he wrote.
Particle provides technology demos via its website for logged out users. There, the article is listed with a summary, a timestamp of when it was last updated, and the source cited in a small section at the bottom.
These sources come from across the political spectrum and include the New York Times, CNBC, AP, ABC, CNN, Breitbart, Guardian, Washington Post, Politico, Fox News, USA Today, Daily Caller, New York Post, The. Hill et al. As the demonstrations have shown, international media outlets are also brought in when necessary. However, each bullet point is not linked to the original source, making it difficult to fact-check the accuracy of the AI summaries without digging into every article. (Key terms are linked, however.) We also noted that the photos attached to the news summaries were watermarked with the publisher's logo.
Image credits: particle
The final product will be different, given that Particle is currently launching a private beta for testing and is hiring a senior iOS engineer, so the company plans to offer a mobile app in the future. There is a high possibility that it will.
A similar model of leveraging a variety of news sources and using AI to summarize them was recently employed by Instagram co-founders' now-shuttered startup Artifact. In these cases, the Artifact team has pre-selected news sources based on factors related to their integrity and quality. For example, dealers had to quickly correct mistakes and be transparent about financing. We look forward to sharing more details about how Particle vets sources closer to public availability.
Bulletin, another AI-powered news app, was also recently launched to provide news summaries and tackle clickbait.
Given the interest in this space, what may set Particle apart is its founding team. Our co-founders, who come from Twitter, have experience of what the real-time news ecosystem is like and have the technical and product experience to build quality products. But it remains to be seen whether publishers who feel AI is encroaching on their space will feel they are being “fairly compensated.”
Adverb Ventures co-founder and managing director April Underwood praised Particle in a LinkedIn post about the company's investment.
“We had the opportunity to help them just as we were completing the first close of Fund 1. We had to wait until the first capital call hit before we could transfer the funds to them. “We didn't,” she said Sunday, adding that Adverb closed its $75 million Fund I just a few months ago. “Sarah and Marcel are the kind of founders we dreamed of supporting when we were trying to start a new early-stage company. They are going after big problem areas. have the skills to tackle big problems in product quality. And they attract other talented people to join them and together invent a future that consumers don't know they (yet) want. “You can,” Underwood wrote.
In an email with TechCrunch, Underwood explained the opportunities ahead:
When it comes to space, we believe AI will impact every aspect of people's digital lives at work and at home. Combine this with the existing conditions at play here. Finding the latest news from trusted sources can be difficult, and the social media landscape is rapidly evolving. And we have to believe that the way people consume news will be different in the future. A few years from now. Sarah and Marcel are uniquely qualified to help people get the news they need in a modern way.
Beykpour told TechCrunch that the idea for Particle came about because there are a lot of challenges around how people get their news and stay updated on what's going on. But the company is still talking with publishers about what they need to feel like they'll be fairly compensated for the model that AI uses to summarize their work.
“To be honest, we're considering it. We're currently talking and working with publishers to find out what the right model is,” she says. “But my goal here is to be fair.”
The company expects to announce more news in this regard in the coming months. In the meantime, Particle's beta sign-up form can be found here.
First publication date is February 26, 2024. Updated with comment from Underwood after publication. 2/26/24, 4:30 PM ET. Funding details updated on February 29th at 8:30am Pacific Time.