Former Twitter engineers are developing Particle, an AI-powered news reader

Former Twitter engineers are developing Particle, an AI-powered news reader

Photo of author

By Vivek Kumar

Former Twitter engineers are developing Particle, an AI-powered news reader

A team headed by the former Twitter engineers are developing Particle, an AI-powered news reader , which can assist people in processing information and news., which became accessible in private beta over the weekend, is a recently launched startup that offers an personalized news-reading experience characterized by a variety of perspectives, all achieved without depending only on AI  to provide a summary of the news however, it also hopes to do it in a manner that is fair rewards publishers and authors in fact, that’s what they claim..

Although Particle hasn’t yet revealed the business plan, it’s coming in a moment when there’s an increasing concern about the effects of AI on the fast decreasing information environment. Information that’s summarized by AI can limit the number of clicks on publishers websites, which could limit the ability of publishers to earn money through advertising could also be diminished.

The company was launched in 2015 by the an earlier Senior Director of Product Management from Twitter, Sara Beykpour who was working on products such as Twitter Blue, Twitter Video as well as conversations. She also was the founder of the app that was in development called Twttr. She worked working at Twitter from 2015 to 2021, advancing her position in software engineering, to the high-level director in product management. Her co-founder was an engineer who was formerly a senior engineer at the two companies, Twitter as well as Tesla, Marcel Molina.

The idea behind Particle that Beykpour stated in the last month is to make it easier for people to keep up with the latest news through AI.

“Sometimes it can feel that headlines are all we are able to do. We want to know more, but quicker,” she wrote in an introduction to the company’s website on the website Threads. “We’re just beginning the process of making use of AI to change how we interact in news.”

With Particle, newsreaders get a short and concise summary of the story with details sourced from a variety of sources. When she announced the beta’s private version, Beykpour noted that readers could utilize the summary to get to speed, or opt to explore the story further to “learn about the way a story unfolds in the course of the course of.”

The venture-backed startup has received money through Kindred Ventures and Adverb Ventures and a variety of angel investors, such as Twitter along with Medium Co-Founder Ev Williams, as well as Behance founder Scott Belsky.

The comment of Belsky on Belsky’s X, “Particle has become an everyday app for me. It takes the various reports (and angles) about any news subject highlights the main aspects as objectively as it is and lets you go deeper into many dimensions. In the age of abstraction, this is a great examples of daily AI,” he wrote.

Particle provides a demonstration of its technology to logged-out users through their website on which articles are highlighted with their description along with the date they last updated, and in a tiny section on the bottom that list the sources they draw from.

These sources come from all over the political spectrum. They include major publishers such as The New York Times, CNBC and the AP, ABC, CNN, Breitbart, The Guardian, The Washington Post, Politico, Fox News, USA Today, The Daily Caller, New York Post, The Hill, and many more. International publications are also pulled from time to time, as the demos show. However each bullet point isn’t linked to the original source, or to any other sources. This means it is difficult to check its accuracy. AI summary without going through every article. (Key terms are however linked). We also noticed that the image that is accompanied by a news report is watermarked by the logo of the publisher.


Images Credits Particle

The final product is likely to differ from the prototype, as Particle is only unveiling its private beta for testing. The company plans to develop a mobile app in the near future, and they are hiring an experienced Chief iOS Engineer.

A similar approach of using various news sources and using AI to summarise it, was used recently by Artifact which is now a closed startup of Instagram’s cofounders. In this instance, the Artifact team was able to select the sources prior to publication by analyzing factors that correlated to their credibility and quality. In particular, the news outlet must be quick to correct mistakes and also be open regarding their sources of funding. We’re looking forward to discussing more detail on the way Particle checks its sources prior to a launch date.

Another AI-powered news application, Bulletin, also recently launched to fight clickbait with providing news summary.

In the current market area, what makes Particle distinguish itself is the team that founded it. After joining Twitter the co-founders have been through the reality of what a real-time media ecosystem is like, and they have the product and technical expertise to develop a top-quality product. The question is whether or not publishers who feel AI is affecting their business will be “fairly compensated” but that remains to be determined.

Adverb Ventures co-founder and Managing Direct April Underwood praised Particle in an update on LinkedIn regarding the company’s acquisition.

“We were able to support them right when we were closing our first close of Fund 1 We needed to wait for our capital call to go through to send them the money and they did!” she said on Sunday and added that Adverb completed its 75 million fund I just a few months ago. “Sara as well as Marcel are the type of founders that we dreamed of backing when we began to start a new, early-stage business. They’re targeting the big problems. They’re equipped to solve big issues at a high standard of quality. And they’re able to attract talented people to join them and work together to create a product that product that consumers aren’t aware of asking about (yet),” Underwood wrote.

Comments on the form weren’t answered. The Particle beta sign-up form can be found available here.