Goal is in a race to develop a more powerful chatbot that resembles OpenAI’s GPT-4, according to a report by The Wall Street Journal. The company plans to start training this new language model in early 2024, and Meta CEO Mark Zuckerberg is pushing to make it free for companies that want to use it in creating artificial intelligence tools.
To reach this goal, Meta has been acquiring AI training chips, specifically the Nvidia H100and building its own data centers so as not to rely exclusively on Microsoft’s Azure cloud platform to train the new chatbot.
In accordance with The Vergethe company assembled a team earlier this year to develop this new model, with the intention of accelerating the creation of artificial intelligence tools capable of emulating human expressions.
An Instagram chatbot?
This goal aligns with the generative AI features Meta is rumored to already be developing, like an Instagram chatbot with 30 different personalitieswhich could be an extension of the AI “people” the company plans to launch soon.
However, Meta has faced significant challenges in its pursuit of advanced AI. Has experienced a turnover of AI researchers this year due to competition for computing resources between multiple large language model (LLM) projects.. Despite the competition, Meta is determined to continue advancing its chatbot development and compete in the generative AI space.
Although OpenAI stated that it was not working on a GPT-5 and would not do so in the near future, other companies such as Apple, Google and Microsoft are also investing in advanced AI models. This indicates that generative artificial intelligence is a constantly evolving field with great potential for applications in a variety of industries.