TIME.CO, Jakarta – Half they used training chips TO THE and build data centers to create new, more powerful chatbots that should be equally sophisticated GPT-4 OpenAI, from the Wall Street Journal menu.
The company reportedly plans to start training a new large language model as early as 2024, with CEO Mark Zuckerberg apparently pushing for the model to give companies the freedom to build AI tools.
The WSJ wrote earlier this week that “Meta purchased more Nvidia H100 AI training chips and upgraded its infrastructure so that, this time, Meta won’t have to rely on Microsoft’s Azure cloud platform to train new chatbots “.
The company reportedly formed a team earlier this year to build the model, with the goal of accelerating the creation of AI tools that can mimic human expressions.
These goals seem like a natural extension of the generative AI features Meta is said to be working on. A June leak claimed that an Instagram chatbot with 30 personalities was being tested, which looked very similar to the previously unannounced AI “characters” the company said it would launch this month.
Announcement
Meta has reportedly dealt with a major turnover of AI researchers this year on sharing computing resources between different LLM projects. Companies also face stiff competition in the field of generative artificial intelligence.
OpenAI said in April that it will not train GPT-5 and “won’t for some time,” but Apple is reportedly spending millions of dollars a day on its “Ajax” AI model that is apparently considered more powerful than GPT-5. 4.
Both Google and Microsoft have expanded the use of AI in their productivity tools, and Google is looking to use generative AI in Google Assistant. Amazon is also undertaking a generative AI initiative across its organization that could lead to chatbot-based Alexa.
Always update latest information. Listen latest news and news selected by Tempo.co on the Telegram channel “Tempo.co Update”. Click https://t.me/tempodotcoupdate join. You have to-to install First the Telegram application.
Quoted From Many Source
