The information: Meta goes all in on open-source AI. The corporate has unveiled LLaMA 2, its first giant language mannequin that’s out there for anybody to make use of—without spending a dime. It’s additionally releasing a model of the AI mannequin that folks can construct into ChatGPT-style chatbots.
Why it issues: The thought is that by releasing the mannequin into the wild and letting builders and firms tinker with it, Meta will be taught vital classes about how one can make its fashions safer, much less biased, and extra environment friendly.
However… Many caveats nonetheless stay. Meta will not be releasing details about the information set that it used to coach LLaMA 2, and it nonetheless spews offensive, dangerous, and in any other case problematic language, identical to rival fashions. Meta additionally can’t assure that it didn’t embody copyrighted works or private knowledge, in keeping with an organization analysis paper shared solely with MIT Expertise Overview. Learn the complete story.
—Melissa Heikkilä
Recognizing Chinese language state media social accounts continues to be a problem
It’s no secret that Chinese language state-owned media are energetic on Western social platforms. However typically they take a covert strategy and distance themselves from China, maybe to succeed in extra unsuspecting audiences.
Such operations have been discovered to focus on Chinese language- and English-speaking customers prior to now. Now, a examine printed final week has found one other community of Twitter accounts that appears to be obscuring its China ties—and this time, it’s made up of Spanish-language information accounts focusing on Latin America. Learn the complete story.
—Zeyi Yang