However, Dr LeCun said last month that Meta has been hampered in part by its reputation as a corporate titan that helps spread lies. Accountable to billions of users, it cannot have a chatbot online that can generate false and biased information.
Spread of misinformation and lies
- deep face swap rule: In most parts of the world, authorities are powerless against deepfakes because few laws exist to regulate the technology. China wants to be the exception.
- Lessons for a new generation: Finland is testing new ways of teaching publicity to students. Here’s what other countries can learn from its success.
- corona myth: Experts say the spread of coronavirus misinformation — especially on far-right platforms like Gab — is likely to be a lasting legacy of the pandemic.no easy solution
- “Talent War”: Seeing misinformation as a potentially costly liability, some companies are trying to hire ex-Twitterers with expertise to rein it in.
“OpenAI and other smaller companies are in a better position to actually get some credit for releasing something like this,” said Chirag Shah, a University of Washington professor who has explored the flaws in technologies like Galactica and ChatGPT. “They’re not going to get the same backlash.”
In recent years, Meta has also shifted its focus to another area of technology: the immersive online worlds of the so-called metaverse, which Mr. Zuckerberg says he sees as the next big thing. In the short term, it’s unclear how the company will offer generative AI offerings through its existing services in a way that will actually capture the public’s attention.
That doesn’t mean it didn’t try. Irina Kofman, senior director of generative AI product management who oversees XAI, said Meta is working quickly to get AI-powered products into the hands of customers. Mr. Zuckerberg is directly involved in steering the initiatives, holding weekly meetings with product heads and top AI researchers, she said.
On a conference call with investors last week, Zuckerberg mentioned artificial intelligence several times, calling it “the foundation of our discovery engine and advertising business,” adding that it will “enable many New products and additional shifts.”
Some Meta executives spoke of the generative AI boom with a decidedly sour grape flavor. In an online discussion last month, Dr. LeCun described ChatGPT as “not particularly innovative” and “not revolutionary” because it relies on technology developed and deployed by Meta, Google and others.
Earlier last year, Meta released BlenderBot, a chatbot that expands on state-of-the-art technology, Dr. LeCun said. But it never caught on, he said, because the company worked hard to make sure it didn’t produce objectionable material.
