Farewell, copy editors? AI is now summarising news on curation app Inshorts 

News curation app Inshorts has started using artificial intelligence (AI) to crunch articles into a 60-word summary, The Economic Times reported.

The Noida-based platform is known for providing short summaries of news stories to its users across categories and these articles are typically edited manually. 

But now, Inshorts chief executive Azhar Iqubal says the company has developed an algorithm called Rapid 60 that can read a news story and condense it, apart from adding a picture and a headline.

He added that Inshorts was planning to use the new AI to scale up the number of articles offered on its platform from 10,000 every month to 1 lakh, and will operate across news verticals.

The AI engine, which was developed over the course of a year, has been trained on more than 5 lakh articles that Inshorts produced with the help of human editors over the past five years.

Iqubal told the newspaper that as opposed to human employees, the AI would not get tired. 

However, Inshorts is not bidding farewell to human editors just yet.

"The top-read stories on the app will still be manually written since they need extreme attention," Iqubal was quoted as saying.

Inshorts was started in 2013 by by Deepit Purkayastha, an IIT Kharagpur dropout, along with IIT Delhi dropouts Iqubal and Arunay Pandey. It has raised more than $25 million in funding so far and is backed by New York-based investment firm Tiger Global as well as Flipkart founders Sachin Bansal and Binny Bansal.

Another domestic content startup NewsBytes, run by Elysium Labs Pvt. Ltd, had told TechCircle earlier this year that it had developed a bot called Yantra which uses machine learning to provide contextual news to readers. 

The team at NewsBytes feeds Yantra input on what news to trawl the Internet for. The information is then edited and presented in a question-and-answer format on an app and website.

“Yantra is based on our in-house AI engine 'Brahmastra'. The engine uses natural language processing techniques and has developed an inverted 'Question-Answering' system where we have fixed questions and a variable set of answers,” Shikha Chaudhry, co-founder of NewsBytes, had said.