Technology

Posted on February 16, 2018 by staff

AI platform fighting fake news could be future of journalism

Technology

A Cambridge University student is looking to solve the problem of fake news with a machine learning platform developed alongside some of the tech world’s finest minds.

Lyric Jain, 21, founded Logically a year ago and says that platforms such as Facebook incentivise clickbait material, which is partly what has led to the rise of fake news.

As has been widely reported in the media, there may also have been Russian political motivations around the Brexit debate and US election.

“A large subset of fake news is just propaganda but there’s also a large subset which is opportunistic,” he told BusinessCloud.

“If someone wants to make money, they don’t care about whether a report is fake and which political spectrum it promotes – they just want the clickbait.

“There needs to be some change in the way that these platforms [such as Facebook] are monetised. Having this clickbait material incentivised isn’t healthy for them.”

The engineering student, who will graduate this year, started his first business at the age of just 16 when he designed and later developed the ABS Secure Lock which is now seen as the industry benchmark for door locks.

His original motivation for building Logically was witnessing the ‘echo chamber’ effect during the Brexit debate where many people were only exposed to opinions which mirrored their own through platforms such as Twitter.

He designed the machine-learning algorithm with the aid of faculty advisers at the Massachusetts Institute of Technology (MIT) in the United States whilst studying at the world-renowned college for a year.

“It works in the same way that a human would go about the process of identifying fake news,” he explained.

“They would look at where the news has been published; who the author is; whether they have any expertise in that field; whether they refer to any primary sources; and ask if the language they’re using is too emotive.

“It cross-checks against official statements and other news outlets and tries to find articles which might be slightly different or even contradictory. Those articles would then be juxtaposed on our platform – users can choose which one to read while their logical accuracy would also be assessed.

“It can determine whether the argument they are using is coherent and come up with a final prediction of how accurate a piece of information is likely to be.”

The platform is also capable of assessing how positive the general sentiment is around a particular topic in the media – and how that has altered over time (see below).

Logically has a board of advisers comprised of alumni from both MIT and Cambridge University and has raised £1 million in funding. It employs 38 people across the UK, India and US and is planning to almost double that in 2018.

The platform – which is planned for a Beta launch an April 2nd and full launch by the end of June – serves as a news aggregator as well as an indicator of factual accuracy.

“It could be the future of journalism,” said Jain, who was brought up in India before moving to the UK when he was 11.

“These social media platforms control what people see and don’t see. Right now we’re quite trusting that Facebook, [its founder] Mark Zuckerberg and Twitter will do the right thing in the public interest. But how long can that continue?

“The easy way of getting rid of fake news would be for Facebook and Twitter to censor everything too aggressively, which has freedom of speech implications.

“We need to come up with a solution but it’s hard to see these companies adopting one standard [of censorship] and applying that internationally.

“We need to educate the platform users that they are exposed to this fake news and are bubbled off in terms of the content they are seeing.

“In some cases you can see some really drastic outcomes as people are exposed to this echo chamber effect: on a countrywide level we’ll end up with more polarised groups of people who don’t speak to each other and have their own version of what’s going on.”