Public, Facebook seems to say that he removes more than 90 percent of hate speech on its platform, but communicating with ordinary people within the company says the figure is alarming at 3% to 5%. This repeats itself: Facebook wants us to believe that almost all hate speech is discarded, when in fact everything is on the platform.
The heinous fraud was exposed amidst a number of complaints, citing thousands of pages of documents that came out, which Facebook’s whistleblower turned Frances Haugen and his legal team filed in the SEC earlier this month. While the public’s interest in the loss focuses on Instagram’s impact on young people’s health (which is not the case with the gun) and the role of News Feed in promoting falsehoods (not revelations), Facebook’s complete failure to reduce hate speech and simple deception. hide this failure is amazing. It reveals how Facebook relies on AI on a small scale, how useless AI is, and the need to force Facebook to come.
In a testimony to the U.S. Congress in October 2020, Mark Zuckerberg described the company’s visibility, which he said shows that “we are well aware, I think, about 94% of the hate speech we have completed.” Witnessing at the House a few months later, Mr. Zuckerberg also answered questions about hate speech by citing a clear report: “We also removed about 12 million pieces of gangs for violating our hate laws, 87% we got by force.” , Facebook announces a number of hate speech in the 80’s and 90’s like this. However, a document released since March 2021 states, “We can take action up to 3-5% of hate… on Facebook.”
Did Facebook really catch a big lie? Yes and no. Ideally, both numbers are correct — they just test different things. The most sought after measure is the one that Facebook has hidden. The measure that Facebook has been publicizing is irrelevant. It’s like every time a police officer pulls you in and asks you how you’re doing, you always answer without ignoring the question instead you brag about the fuel flow of your car.
There are two ways in which rude words can be put in place to be reviewed and removed. Users can specify it manually, or AI algorithms can try to identify it easily. Algorithmic recognition is important not only because it is practical, but also because it is exhausting strongly, before anyone else could utter an insult.
The 94% publicity that Facebook has publicly declared is “the amount of what is going on,” the amount of hate speech removed that Facebook AI recognized, shared by the amount of rude words that were removed. It is the amount of hate speech captures that are removed before a user says them.
Facebook may want you to think that this statistic reflects the amount of hate speech that is removed before it has a chance to hurt — but all that is tested is how much of a factor is the recognition of language on the platform.
What matters to the community is the amount of hate speech that exists no removed from the platform. The best way to record this is to count the number of hate speech images that have been distributed in all forms of hate speech. This “downturn” measures the amount of hate speech on Facebook that has been dropped – and that is the number that Facebook has tried to hide.
Thanks to Haugen, we finally know the amount, and it’s a pity. According to internal records, more than 95% of hate speech shared on Facebook resides on Facebook. Zuckerberg boasted to Congress that Facebook had removed 12 million hatreds from groups, but based on the estimates released, we now know that about 250 million pieces of hatred may have remained. This is shocking, and shows the decline in performance that has taken place since the early days of unofficial internet circles — despite the huge amount of money that Facebook has been doing in favor of AI over the years.
Unfortunately, Haugen’s legal team sued the SEC for interfering in the case by boldly asserting, “Facebook Post Proves Facebook’s False Statement.” That alone is a lie: Facebook’s reliance on high prices is a bad test to say. Facebook has not lied or “distorted” the truth, as the allegations claim — but has repeatedly and unequivocally exposed the cesspool of hate speech on its platform, and the threat the company is having to resume.