Meta’s new AI tool Few-Shot Learner can detect error messages with fewer or no instances

Meta recently announced a new AI tool called Few-Shot Learner, which has been used on its platforms such as Facebook to counter misinformation and other harmful content. With Few-Shot Learner, Meta’s system can adapt to new threats faster and requires fewer content-sometimes even with zero marked examples to learn.

In addition to speeding up the learning process of AI, Few-Shot Learner AI can span more than 100 languages, which means that some COVID-19 error messages in French will be translated into another language such as English and allowed to be in English. Was stopped before raging in the conspiracy circle. Like some other AI systems in Meta, Few-Shot Learner can learn from different kinds of data, including images and text.

Join RealMi Central on Telegram, Facebook & Twitter

When discussing the potential of this new technology, Meta said: We believe that over time, FSL can improve the performance of all our honest AI systems, allowing them to use a single, shared knowledge base and backbone to handle many different types. Violations. Although there is still a lot of work to be done, these early production results are an important milestone that heralds the shift to smarter and more versatile AI systems.

Meta said that from training to deployment, FSL can be deployed in about six weeks, while traditional AI systems designed for the same task can take up to six months to complete deployment, which makes it easier for misinformation and harmful content to spread.

Leave a Comment