{"id":49143,"date":"2022-08-11T01:18:12","date_gmt":"2022-08-11T01:18:12","guid":{"rendered":"https:\/\/tahrir2day.com\/news\/2022\/08\/11\/metas-chatbot-says-the-company-exploits-people\/"},"modified":"2022-08-11T01:18:12","modified_gmt":"2022-08-11T01:18:12","slug":"metas-chatbot-says-the-company-exploits-people","status":"publish","type":"post","link":"https:\/\/tahrir2day.com\/news\/2022\/08\/11\/metas-chatbot-says-the-company-exploits-people\/","title":{"rendered":"Meta’s chatbot says the company ‘exploits people’"},"content":{"rendered":"
\n

Meta’s new chatbot prototype has told the BBC that Mark Zuckerberg is exploiting his users for money. <\/b><\/p>\n<\/div>\n

\n

According to Meta, the chatbot uses artificial intelligence and can chat about “almost any topic.” <\/p>\n<\/div>\n

\n

When asked what the chatbot thought of the company’s CEO and founder, he replied, “Our country is divided and he hasn’t helped it at all.” <\/p>\n<\/div>\n

\n

Meta said the chatbot is a prototype and could provide rude or offensive responses. <\/p>\n<\/div>\n

\n

The chatbot, dubbed BlenderBot 3, was released to the public on Friday.<\/p>\n<\/div>\n

\n

The program “learns” from large amounts of publicly available language data. <\/p>\n<\/div>\n

\n

When asked about Mark Zuckerberg, the BBC’s chatbot said: “He did a terrible job testifying before Congress. It worries me for our country.\u201d <\/p>\n<\/div>\n

\n

Mr. Zuckerberg has been questioned by US politicians on a number of occasions, most notably in 2018. <\/p>\n<\/div>\n

\n

“Our country is divided and he didn’t help at all,” the chatbot continued.<\/p>\n<\/div>\n

\n

“His company exploits people for money and he doesn’t care. It has to stop!\u201d it said.<\/p>\n<\/div>\n

\n

Meta has been criticized for not doing enough to prevent the spread of disinformation and hate speech on its platforms. Last year, a former employee, Frances Haugen, accused the company of putting profits ahead of online safety. <\/p>\n<\/div>\n

\n

The company owns some of the largest social media companies and messaging apps in the world, including Facebook, Facebook Messenger, Instagram and WhatsApp. <\/p>\n<\/div>\n

\n

BlenderBot 3’s algorithm searches the internet for answers. It’s likely that his views on Mr. Zuckerberg were “learned” from other people’s opinions that the algorithm analyzed. <\/p>\n<\/div>\n

\n

The Wall Street Journal has reported<\/a> BlenderBot 3 told one of its journalists that Donald Trump was and always will be the US President. <\/p>\n<\/div>\n

\n

A Business Insider journalist said the chatbot called Mr Zuckerberg “creepy”. <\/p>\n<\/div>\n

\n

Meta made the BlenderBot 3 public and risked bad publicity for good reason. It needs data. <\/p>\n<\/div>\n

\n

“Allowing an AI system to interact with humans in the real world leads to longer, more diverse conversations, as well as more diverse feedback,” Meta said in a blog post. <\/p>\n<\/div>\n

\n

Chatbots that learn from human interactions can learn from their good and bad behavior.<\/p>\n<\/div>\n

\n

In 2016, Microsoft apologized after Twitter users taught its chatbot to be racist.<\/p>\n<\/div>\n

\n

Meta accepts that BlenderBot 3 can say the wrong thing – and imitate language that could be “unsafe, biased, or offensive”. The company said it installed security measures, but the chatbot could still be rude.<\/p>\n<\/div>\n

\n

When I asked the BlenderBot 3 what it thought of me, it said it had never heard of me. <\/p>\n<\/div>\n

\n

“He must not be so popular,” it said. <\/p>\n<\/div>\n

\n

Harsh. <\/p>\n<\/div>\n

\n

James Clayton is the BBC’s North American technology reporter based in San Francisco. Follow him on Twitter <\/i>@jamesclayton5<\/a>.<\/p>\n<\/div>\n