Browsed by
Month: March 2019

Facebook’s false positive censorship?

Facebook’s false positive censorship?

This image was censored by Facebook. Why? No one knows. But I suspect it was due to an image to text conversion error, or to AI language analysis that misinterpreted the meaning of the text. AI has been promised as the savior of social media but in reality, it continues to have numerous problems.

Twitter admits it is shadow banning tweets

Twitter admits it is shadow banning tweets

Shadow banning is common on social media, including newspaper comments. Rather than outright ban a comment, the media publisher hides the comment from everyone except the person who wrote it. In this way, the comment author is unaware that they have been censored.

Nice explanation of how propaganda messaging works

Nice explanation of how propaganda messaging works

“A good cause makes a lie easier to buy”. This is why many propaganda campaigns use themes such as “for the children” – even though their claims are exaggerated or not true. The target buys the propaganda message since its for a good cause.

Fake reviews now being written by bad AI software

Fake reviews now being written by bad AI software

Fake news, fake reviews – its all the same after a while. Any web service relying on user contributed content has devolved into the noise of fake posts. Now, AI is being applied to writing fake reviews, in an attempt to get around the AI code that seeks to detect fake reviews.