Meta is assembling AI training chips and building data centers to build a new, more powerful chatbot that it hopes will be as sophisticated as OpenAI’s GPT-4. It is stated that the company plans to start training its new large language model in early 2024, and CEO Mark Zuckerberg wants this model to be free once again for companies to create artificial intelligence tools.Free AI powered by GPT-4
Meta is purchasing more Nvidia H100 artificial intelligence training chips to create its free artificial intelligence, which is as functional as the GPT-4 model of OpenAI, one of the leading companies in artificial intelligence. Meta will not have to rely on Microsoft’s Azure cloud platform by strengthening its infrastructure with new H100 solutions. The company also reportedly formed a group to build the model earlier this year, with the aim of accelerating the creation of AI tools that can mimic human expressions. So the foundations have already been laid.
This goal seems like a natural extension of the generative AI features Meta is said to be currently working on. A leak in June claimed that an Instagram chatbot with 30 personalities was being tested, which is very similar to the unannounced AI “personas” the company is rumored to be releasing this month.
Meta has implemented multiple LLM projects this year, many of which are free and open source. Although Meta invests in the field of productive artificial intelligence, other companies are not idle either. Microsoft and Google have been working in this field for quite some time already. Recently, it was reported that Apple is pouring millions of dollars every day into a model codenamed “Ajax”, which is said to be even more powerful than GPT-4.
58343
