Topics
Latest
AI
Amazon
Image Credits:Particle
Apps
Biotech & Health
Climate
Image Credits:Particle
Cloud Computing
Commerce
Crypto
Image Credits:Particle
Enterprise
EVs
Fintech
Image Credits:Particle
fundraise
contrivance
stake
Image Credits:Particle
Government & Policy
Hardware
Layoffs
Media & Entertainment
Meta
Microsoft
Privacy
Robotics
Security
Social
Space
startup
TikTok
Transportation
Venture
More from TechCrunch
Events
Startup Battlefield
StrictlyVC
Podcasts
video
Partner Content
TechCrunch Brand Studio
Crunchboard
meet Us
The media diligence today may not have a very favorable view of AI — a engineering science that ’s already been used to replace newsman withAI - written copy , while other AI party havescooped up journalists ’ workto feed their chatbots ’ data demands , but withoutreturningtraffic to the publishing house as search railway locomotive once did . However , one startup , an AI newsreader calledParticlefromformer Twitter locomotive engineer , believes that AI could wait on a valuable function in the medium industry by help consumer make sense of the newsworthiness and delve deeply into level , while still finding a way of life to abide the publishers ’ business .
back by$4.4 millionin seed funding and a$10.9 million Series Aled by Lightspeed , Particle was set up last twelvemonth by the former aged director of Product Management at Twitter , Sara Beykpour , who worked on ware like Twitter Blue , Twitter Video , and conversations , and who spearheaded theexperimental app , twttr . Her co - founding father is a former senior railroad engineer at Twitter and Tesla , Marcel Molina .
From the consumer ’ perspective , the core group idea behindParticleis to help reader better understand the newsworthiness with the help of AI engineering science . More than just summarizing stories into key bullet points for nimble snatch - ups , Particle offer a kind of ingenious features that let you draw near the news in different fashion .
But instead of plainly suckle up publishers ’ work for its own use , Particle aims to compensate publisher or even drive traffic back to news sites by prominently showcasing and colligate to reference instantly underneath its AI summary .
To start , Particle has partner with specific newspaper publisher to host some of their content in the app via their APIs , including outlets like Reuters , AFP , and Fortune . These better half receive serious placement and their links are play up in Au above others .
Join us at TechCrunch Sessions: AI
Exhibit at TechCrunch Sessions: AI
Already , beta exam indicate that proofreader are clicking through to publishers ’ sites because of the app ’s design and user user interface , though that could tilt now that the app is launching beyond news junkies to the general world . In clip , the company intends to introduce other ways to lick with the sensitive , too , in addition to ship them referral traffic . The squad is also having discussions with publisher about providing its users access to paywalled cognitive content in a way that cause sense for all parties .
“ get bass partnerships and collaboration is one of the thing that we ’re really interested in , ” mention Beykpour .
To avail with its traffic referral sweat , the app ’s article section includes self-aggrandising pat targets , attain it easy for lector to click through to the publishing house ’s web site . Plus , Particle admit the faces of the journalists on their bylines , and lector can come after through links to publisher profiles to say more of their subject or play along them .
Using the app ’s construct - in AI tools , news consumers can exchange between different modes like “ explicate Like I ’m 5 , ” to get a simplified version of a complicated story or those that summarize “ just the facts , ” ( or the 5W ’s — who , what , when , where , and why ) . you could have the news summarized in another language besides English , or listen to an audio sum-up of a tale or a personalized selection of stories while on the go . mote can also displume out important quote from a taradiddle and other links of reference .
But two of the more interesting features involve how Particle leverage AI to serve present the intelligence from different angles and allows you to further engage with the chronicle at hand by asking questions .
In Particle , one tool call in “ Opposite Sides ” aims to snap off exploiter ’ filter bubble by present different viewpoints from the same tale . This poser has been tried before by other news apps , includingthe startup BriefandSmartNews . Unlike early try , Particle includes a story spectrum that show how news is being reported across both “ red-faced ” and “ blue”-leaning site , with bubble placed to indicate how far to the left or right the news ’ positioning is , and how outsized the coverage may be from one side or the other . The AI will also summarize both incline ’ positions , allowing news show consumers to reach their own feeling about the matter .
However , the app ’s killer feature is an AI chatbot that have you ask doubtfulness and get instant answers about a story . The app will let in suggested doubt and those asked by others . For model , if you ’re take about Trump ’s immigration insurance policy plans , you could ask the chatbot thing like “ What are the potential effectual challenges to Trump ’s exile plans ? ” or “ What are the possible cost of mass deportation ? ” among other things . mote will then use its AI applied science to witness those answers and fact - check them for truth .
“ The Old World chat function uses OpenAI as well as … our own pre - processing and post - processing , ” explains Beykpour , in an interview with TechCrunch . “ It uses the contentedness , search the vane a little bit — if it want to find extra information on the web — to generate those answers . ” She says that after the answer is render , Particle includes an extra whole tone where the AI has to go find the supporting material that matches those answers .
Overall , the app encompasses technical school like OpenAI ’s GPT-4o and GPT-4o mini , Anthropic , Cohere , and others , include more traditional AI technologies , which are not LLM - base , from Google .
“ We have a processing word of mouth that takes related subject matter and sum up it into fastball points , into a headline , sub - newspaper headline , and does all the descent , ” she continues . “ Then … we pull out quotes and link and all sorts of relevant information about [ the story ] . And we have our own algorithms to outrank , so that the most important or relevant connection 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 ship’s company claims that its engineering science cut back AI accuracy problems that would otherwise occur one out of 100 times , and reduce their likeliness to one out of 10,000 times .
Particle will also use human editor program as it grows to help substantially manage the AI mental object and curate its home page , she note .
The app isa innocent download on iOSfor the time being and work across iPhone and iPad .
