Kicking off this week’s Friday Social with a bit of news straight from the future, Facebook has revealed that it shut down two of its artificial intelligence robots after they invented their own language.
The bots were created by Facebook’s Artificial Intelligence Research team, and were intended to learn how to negotiate by mimicking human bartering.
However, the robots began to learn their own form of communication – ‘divergence from human language’ – and were soon killed off.
Dhruv Batra, a Facebook researcher, said it wouldn’t be possible to translate the language back into English: “It’s important to remember, there aren’t bilingual speakers of AI and human languages.”
In other AI news, Google is improving its efforts to fight extremist content on YouTube with machine learning.
In a blog post, the Google team updated on ‘better detection and faster removal’ being driven by AI. Progress listed includes:
Speed and efficiency: Our machine learning systems are faster and more effective than ever before. Over 75 percent of the videos we’ve removed for violent extremism over the past month were taken down before receiving a single human flag.
Accuracy: The accuracy of our systems has improved dramatically due to our machine learning technology. While these tools aren’t perfect, and aren’t right for every setting, in many cases our systems have proven more accurate than humans at flagging videos that need to be removed.
Scale: With over 400 hours of content uploaded to YouTube every minute, finding and taking action on violent extremist content poses a significant challenge. But over the past month, our initial use of machine learning has more than doubled both the number of videos we’ve removed for violent extremism, as well as the rate at which we’ve taken this kind of content down.
The blog post also states that when users search for ‘sensitive keywords’ on YouTube, they’ll instead be redirected to a playlist of content that debunks extremist messages.
Research by the University of Cambridge has unveiled that Twitter accounts with huge follower numbers behave just like everyone’s favourite Twitter bots.
The research was looking at how bots behave, and used an algorithm capable of identifying these with 86% accuracy with factors including account creation date, tweet frequency, replies to tweets, likes received, and more.
According to researchers, accounts with more than 10m followers are very similar to bot accounts with similar numbers.
The research flagged that not all bots are what we think of as the traditionally spammy posters – some are benign, such as news accounts which rely on automation to share content efficiently.
PhD student Zafar Galani said: “Many people tend to think that bots are nefarious or evil, but that’s not true. They can be anything, just like a person. Some of them aren’t exactly legal or moral, but many of them are completely harmless. What I’m doing next is modelling the social cost of these bots – how are they changing the nature and quality of conversations online? What is clear though, is that bots are here to stay.”
Facebook has this week announced that it’ll begin prioritising stories in the News Feed that link to faster-loading pages – including their own Instant Articles.
In a blog post, Jiayi Wen and Shengbo Guo, both Facebook Engineers, explained that user frustration in slow-loading pages is behind the change – apparently, as many as 40% of website visitors abandon a site after just three seconds of delay.
Factors to be taken into account include network connection and the general speed of the corresponding website, and the post explains that most Pages won’t see any significant change. Those with slower-loading websites will see decreases in referral traffic – as our own biddable media analyst Rachel points out, although you could see more prioritisation in the News Feed using Instant Articles, that referral traffic will then disappear altogether, presenting a Catch 22 for brands.
According to reports across social media this Thursday (03 August), short-term White House Director of Communications Anthony Scaramucci will be hosting an ‘online event’ (livestream) on Friday 04 August to ‘tell his side of the story’.
We’re ready and waiting with popcorn, Mooch.
Anthony Scaramucci is hosting an "online event" to tell his side of the story https://t.co/89QB6Eo5Fw
— New York Magazine (@NYMag) August 3, 2017
PR & social media, with a bit more