Meta Seeks To Develop An Ai Model That Performs Better Than Openai’s Gpt-4

Meta is reportedly developing a powerful new chatbot that is expected to be as intelligent as OpenAI’s GPT-4. To do this, the company has purchased AI training chips and is building data centers. Meta has also upgraded its infrastructure and added more Nvidia H100 AI-training processors. Additionally, the company has developed Code Llama, an AI tool for writing new code and fixing human-written work. To further improve the AI model, Meta has installed a reinforcement learning framework, allowing the AI to continuously learn and adapt in real time. Finally, the company is rumored to be releasing AI “personas” later this month and an Instagram chatbot with 30 distinct personalities in June.