Friday, August 23, 2019

Fake News/rev XX


Fake news is everywhere. From Fox News to CNN, spin is part of the vocabulary of the news channels to 1, sell ads to the specific viewer demographic in question and 2, to shape public opinion as the real owners, so eloquently articulated by the late great George Carlin, don't want critical thinkers capable of understanding just what the hell is going on in a world spinning increasingly out of control. With that being said, more fake news is being generated by AI because 1, it's CHEAPER and 2, AI is getting better at it.

AI vs AI

AI can be used to spread fake news, write fake reviews and to create a pretend mob of social media users aimed at bombarding comments sections with specific agendas.

However, according to MIT researchers, it can now also be used to spot fake artificially-generated text — apparently, it takes one to know one.

But there's hope.

Though the technology for misinformation is advancing at a worryingly fast pace, the same broad toolset can thankfully be used to catch this type of misinformation. Fake news, deepfakes and twitter bots might have their days numbered by the very technology that helped create them.

Harvard University and MIT-IBM Watson AI Lab researchers recently developed a new tool that spots text that has been generated by AI.

The tool called the Giant Language Model Test Room (GLTR), takes advantage of the fact that AI text generators use fairly predictable statistical patterns in text.

While these patterns might not be easy to spot for your average reader, it seems that an algorithm can do a pretty good job at it. The AI tool, essentially, can tell if the text is too predictable to have been written by a human.

Factoid:  6 corporations own 90% of all US media.




No comments: