The social media giant Facebook (which changed its name to META) has decided to adopt a new security strategy. It has trained unprecedented metrics and technologies to more easily detect malicious content on the platform.
Keen to deal with the waves of hate, misinformation and harmful content sweeping the platform, Facebook officials have decided to radically change their strategy.
Facebook has put all its weapons to the test
The most innovative AI systems have now been deployed Facebook social networking siteTo succeed in doing its social platform A safer and more enjoyable environment for users who visit and use them.
“Malicious content can evolve rapidly, which is why we have built new artificial intelligence technology that can more easily adapt to work faster on new or changing types of malicious content,” network representatives said. The most popular social networking site in the world.
According to Facebook’s announcement, this new AI system uses “accelerated learning”.
According to Facebook, the AI technology is called Few-Shot Learner (FSL) and it can adapt to new types of content faster. Harmful or on advanced content. The technology “learns” from different types of data and can “enable” AI models already in use to discover different types of content.
According to the explanations of Facebook computer scientists, the method they are currently using is called Simple Learning (FSL) and the company has been testing it since the end of 2020.
Mathematical models start with a general and expanded understanding of many topics, then you need a much smaller number of categorized examples, sometimes even zero, to learn new tasks. The company likens its method to a fishing rod that can catch multiple types of fish versus a simple fishing rod that can catch just one type of fish.
“Certified tv guru. Reader. Professional writer. Avid introvert. Extreme pop culture buff.”