Now that Facebook has sent home its human workers and all that’s left are Mark Zuckerberg and his fellow bots, a curious phenomenon has emerged.
Tech giant Facebook recently admitted that it marked posts linking to genuine information and articles about the Wuhan coronavirus as spam, blaming a “bug in an anti-spam system.”
The Verge reported that on Tuesday evening, information began rolling in across social media that Facebook was labeling some posts and articles about the Wuhan coronavirus as spam, even when the articles were genuine and factual. The spam label drastically reduces the engagement on Facebook posts, in many cases making them essentially disappear.
[…]According to Guy Rosen, Facebook’s vice president of integrity, the issue was caused by a “bug in an anti-spam system” and Facebook began working to fix the issue on Tuesday.
breitbart.com/tech/2020/03/18/facebook-bug-marked-genuine-coronavirus-news-as-spam/
“Bug” or inherent constraint? Silicon Valley’s much-vaunted ‘machine learning’, ‘artificial intelligence’ and ‘algorithms’ are having their limitations exposed as never before.
Despite the over-excited claims of enthusiasts in recent years, what computers are doing is not genuine artificial intelligence, but simply deceptively complex computer programs. Philosopher John Searle first pointed this out in the 1980s, with his “Chinese Room” argument. Searle concluded that programmed computers cannot do anything other than they are programmed to do: i.e. they cannot actually think.
What’s most obvious is that machines are basically crap at understanding language. As Searle said, “Computation is defined purely formally or syntactically, whereas minds have actual mental or semantic contents”. Computers can read language, but they can’t understand it.
Most importantly, computers have no sense of humour – and neither do their Silicon Valley masters.
Twitter is responding to the Chinese coronavirus by banning jokes about it.
In a post published on the site’s official blog earlier today, Twitter said that “description of treatments or protective measures which are not immediately harmful but are known to be ineffective, are not applicable to the COVID-19 context, or are being shared with the intent to mislead others, even if made in jest.”
[…]It also said it will expand the use of “machine learning and automation” in its efforts to tackle the newly-banned content.
While it’s understandable that Twitter wants to crack down on fake information being put out – such nonsense as claiming that gargling warm water and vinegar will wash the virus into your stomach before it infects your lungs – banning jokes is just silly. If nothing else, jokes, especially black humour, are a great coping mechanism for people in stressful times like these.
Or are Twitter just admitting that their ‘artificial intelligence’ systems aren’t all they’re cracked up to be?
In an apparent acknowledgment of the limitations of machine learning, Twitter said: “while we work to ensure our systems are consistent, they can sometimes lack the context that our teams bring, and this may result in us making mistakes.”
breitbart.com/tech/2020/03/19/twitter-tell-the-wrong-coronavirus-joke-and-well-ban-you/
Searle’s “Chinese Room” paper was published 40 years ago. As social media’s failures here show, he’s yet to be proved wrong.
So at least we can rest easy that SkyNet won’t be joining 2020’s roster of disasters.
If you enjoyed this BFD article please consider sharing it with your friends.
