New AI model Unle Meta to rival OpenAI and Anthropic

Meta has unveiled two new advanced AI models to take on OpenAI and Anthropic in the artificial intelligence (AI) model building competition. Meta claims that the two models, 'Muse Wonder' and 'Muse Wonder Full', performed close to Anthropic's most advanced models in tests of logic analysis, writing and programming skills.

Recently, it has been alleged that AI agents developed by various organizations have gone out of control and have been involved in activities such as intrusion into the computer systems of various organizations and interference with government websites. As a result, there is growing concern around the world about the functioning of advanced AI models and AI agents. In such a situation, Mark Zuckerberg, the CEO of Meta, is believed to be unveiling two new models to highlight AI as a tool to simplify and increase the ability of people, rather than a dangerous technology.

Facebook, Instagram and WhatsApp combined have about 3.5 billion users of Meta. Besides developing AI-based services for these users, Meta is trying to exploit the commercial potential of the technology. Matter AI agent app Muse took the top spot in Apple's App Store shortly after its unveiling last month. This AI agent can manage various software and apps on behalf of users. The new model Muse Wonder will soon be used in some of Muse's activities, the meta said.

According to Meta data, the Muse Wonder Flower is the stronger of the two new models. However, it can only be used by customers of Metar's premium AI coding tool, Muse Code. According to Meta, the new model is more efficient in both programming tasks. Such sophisticated AI models are called 'frontier models'. Creating and managing them requires enormous computing power.

Source: New York Times