Meta Advances Development of Artificial General Intelligence (AGI)
In the dynamic world of Artificial Intelligence (AI), Meta, the company behind Facebook, has announced its plans for the further development and integration of its AI research efforts. This marks a crucial step towards the realization of Artificial General Intelligence (AGI), an advanced form of AI capable of handling a broad spectrum of cognitive tasks. The integration of their two main research teams, FAIR (Fundamental AI Research) and GenAI, underscores Meta’s commitment to pushing the boundaries of AI technology.
A key component of Meta’s strategy is the development of Llama 3, a large language model (LLM) known for its ability to process and generate human language. This model is part of the endeavor to achieve AGI, requiring significant advancements in areas such as reasoning, planning, coding, memory, and other cognitive abilities.
To reach its ambitious goals in AI, Meta is heavily investing in its computing infrastructure. By the end of the year, the company plans to acquire about 350,000 Nvidia H100 graphics processors (GPUs), raising the total number of GPUs used for AI training purposes to nearly 600,000. This computing power is essential to meet the complex requirements of AGI training and represents a remarkable demonstration of Meta’s resource strength.
In comparison, OpenAI has used around 20,000 Nvidia GPUs for training its GPT models so far. However, reports suggest that OpenAI aims for a significantly higher number of GPUs for future models, possibly up to 10 million, to develop advanced AGI models. This number, however, has not been officially confirmed. While OpenAI develops proprietary models ensuring controlled quality and safety, Meta adopts an open-source approach, enabling broader collaboration and innovation. This could lead to wider acceptance and application of AI technologies in the long run but also poses challenges regarding safety and potential misuse.
Besides pure software development, Meta is also driving the development of AI-centric hardware devices, as demonstrated by the Ray-Ban Meta Smart Glasses. These devices showcase Meta’s vision of integrating AI and the Metaverse, offering a new interface for AI interaction in daily life.
Mark Zuckerberg’s announcement can be viewed in detail here:
General Note: This text was co-generated with the help of ChatGPT, and/or technical terms were explained by a GPT.