AI That Busts Fake News Will Be First True AI, Experts Suggest

Dec 19, 2016 05:53 AM EST

Artificial intelligence and developments in the field are making staggering revolutionary implications in the world today. So is there room for fake news?

Gadgets are now recognizing English as well as humans, smartphones can instantly translate conversations, and self-driving cars are becoming the norm. 

However, it appears we may be able to tackle fake news with AI as well. According to Wired, Dean Pomerleau can be remembered to have built a self-driving car way back in 1989, using the same technology (neural networks) that power modern gadgets like the Amazon Echo and Microsoft Translator.

His car wasn't ready for highways at the time, given the limited computing power of the time period, but he was one of the few who understood the implications of AI even before we had the technology to understand them.

Now Pomerleau believes that AI may be the solution to tackle fake news. After the recent elections, he has called the AI community to build an algorithm that can identify fake news and remove it from services like Google, Twitter, and Facebook.

According to the Fake News Challenge, he even offers $1,000 to people who want to tackle the task. And there are! Nearly 40 researchers of varying degrees are joining the grassroots project.

Delip Rao, a machine learning expert that helped build the speech recognition system that underpins the Amazon Echo, put in another $1,000. 

However, it appears this is only something we could hope for. According to the New York Times, while neural networks can recognize objects in videos and even spot computer viruses, it will be hard for them to identify fake news with real certainty.

This is because the characteristics of fake news stories are enormous and are hard to pin down. For instance, knowing what's fake doesn't need pattern recognition but human judgment himself.

A machine that can reliably identify fake news is a machine that has completely showed AI.

This means Pomerleau hopes that other researchers can build algorithms to mitigate these problems, that an algorithm can flag "potentially" fake news for humans to review.

This is another case of AI working alongside humans, in ways that help them increase speed and accuracy. 

Of course, this may also mean some companies are also looking into the matter of fake news. For instance, Yann LeCun, the head of AI research at Facebook, told a group of reporters that technology could solve the fake news problem. 

However, it may be a "universal" consensus that the best way to tackle the issue so far is to include humans in the picture.

Wired suggests the best way to do this is to start with an online database with bogus stories, like what Stanford University did to ImageNet. They stocked it with photos, hoping to facilitate computer vision. Now they were able to use it to help their AI recognize objects and similar faces. 

© 2018 All rights reserved. Do not reproduce without permission.
© Copyright 2018 NATURE WORLD NEWS All rights reserved.
About Us Contact Us Privacy Policy Terms&Conditions