The media trade immediately could not have a really favorable view of AI — a know-how that’s already been used to exchange reporters with AI-written copy, whereas different AI firms have scooped up journalists’ work to feed their chatbots’ knowledge calls for, however with out returning visitors to the writer as engines like google as soon as did. Nevertheless, one startup, an AI newsreader known as Particle from former Twitter engineers, believes that AI might serve a helpful position within the media trade by serving to customers make sense of the information and dig deeper into tales, whereas nonetheless discovering a option to assist the publishers’ companies.
Backed by $4.4 million in seed funding and a $10.9 million Collection A led by Lightspeed, Particle was based final 12 months by the previous senior director of Product Administration at Twitter, Sara Beykpour, who labored on merchandise like Twitter Blue, Twitter Video, and conversations, and who spearheaded the experimental app, twttr. Her co-founder is a former senior engineer at Twitter and Tesla, Marcel Molina.
From the customers’ perspective, the core thought behind Particle is to assist readers higher perceive the information with the assistance of AI know-how. Extra than simply summarizing tales into key bullet factors for fast catch-ups, Particle provides quite a lot of intelligent options that allow you to strategy the information in several methods.
However as a substitute of merely sucking up publishers’ work for its personal use, Particle goals to compensate publishers and even drive visitors again to information websites by prominently showcasing and linking to sources straight beneath its AI summaries.
To begin, Particle has partnered with particular publishers to host a few of their content material within the app through their APIs, together with retailers like Reuters, AFP, and Fortune. These companions obtain higher positioning and their hyperlinks are highlighted in gold above others.
Already, beta checks point out that readers are clicking by to publishers’ websites due to the app’s design and consumer interface, although that would shift now that the app is launching past information junkies to most people. In time, the corporate intends to introduce different methods to work with the media, too, along with sending them referral visitors. The group can be having discussions with publishers about offering its customers entry to paywalled content material in a method that is sensible for all events.
“Having deep partnerships and collaboration is one of the things that we’re really interested in,” notes Beykpour.
To assist with its visitors referral efforts, the app’s article part contains large faucet targets, making it straightforward for readers to click on by to the writer’s website. Plus, Particle contains the faces of the journalists on their bylines, and readers can comply with by hyperlinks to writer profiles to learn extra of their content material or comply with them.
Utilizing the app’s built-in AI instruments, information customers can swap between completely different modes like “Explain Like I’m 5,” to get a simplified model of a sophisticated story or people who summarize “just the facts,” (or the 5W’s — who, what, when, the place, and why). You may have the information summarized in one other language in addition to English, or take heed to an audio abstract of a narrative or a personalised collection of tales whereas on the go. Particle also can pull out essential quotes from a narrative and different hyperlinks of reference.
However two of the extra attention-grabbing options contain how Particle leverages AI to assist current the information from completely different angles and means that you can additional interact with the story at hand by asking questions.
In Particle, one instrument known as “Opposite Sides” goals to interrupt customers’ filter bubbles by presenting completely different viewpoints from the identical story. This mannequin has been tried earlier than by different information apps, together with the startup Transient and SmartNews. In contrast to earlier efforts, Particle features a story spectrum that exhibits how information is being reported throughout each “red” and “blue”-leaning websites, with bubbles positioned to point how far to the left or proper the information’ positioning is, and the way outsized the protection could also be from one aspect or the opposite. The AI may also summarize either side’ positions, permitting information customers to succeed in their very own opinions in regards to the matter.
Nevertheless, the app’s killer characteristic is an AI chatbot that permits you to ask questions and get prompt solutions a couple of story. The app will embrace recommended questions and people requested by others. For instance, when you’re studying about Trump’s immigration coverage plans, you possibly can ask the chatbot issues like “What are the potential legal challenges to Trump’s deportation plans?” or “What are the potential costs of mass deportation?” amongst different issues. Particle will then use its AI know-how to search out these solutions and fact-check them for accuracy.
“The chat function uses OpenAI as well as…our own pre-processing and post-processing,” explains Beykpour, in an interview with TechCrunch. “It uses the content, searches the web a little bit — if it wants to find extra information on the web — to generate those answers.” She says that after the reply is generated, Particle contains an additional step the place the AI has to go discover the supporting materials that matches these solutions.
Total, the app encompasses tech like OpenAI’s GPT-4o and GPT-4o mini, Anthropic, Cohere, and others, together with extra conventional AI applied sciences, which aren’t LLM-based, from Google.
“We have a processing pipeline that takes related content and summarizes it into bullet points, into a headline, sub-headline, and does all the extractions,” she continues. “Then…we pull out quotes and links and all sorts of relevant information about [the story]. And we have our own algorithms to rank, so that the most important or relevant link is the one that you see first — or what we think is the most important or relevant quote is the one that you see first.”
The corporate claims that its know-how reduces AI accuracy issues that will in any other case happen one out of 100 occasions, and reduces their probability to 1 out of 10,000 occasions.
Particle may also use human editors because it grows to assist higher handle the AI content material and curate its homepage, she notes.
The app is a free obtain on iOS in the intervening time and works throughout iPhone and iPad.
TechCrunch has an AI-focused e-newsletter! Enroll right here to get it in your inbox each Wednesday.
