Meta Developing Advanced AI
According to a report by the Wall Street Journal, Meta is currently developing a robust AI model with the goal of matching OpenAI’s GPT-4. This new model is slated for training in early 2024 and is expected to surpass the capabilities of Meta’s recently unveiled ‘Llama-2.’
The primary objective of this advanced model is to assist businesses in generating high-quality text, conducting sophisticated analyses, and more. Notably, it will undergo training on Meta’s infrastructure and will be made available as an open-source resource, aligning with Mark Zuckerberg’s commitment to democratizing technology access.
It is worth noting that the release of Meta’s AI model is anticipated to occur subsequent to the launch of Gemini, Google’s LLM, which is rumored to aim for even greater capabilities than GPT-4.
View our content creation services for your business. Blog posts. Social Media posts. Website copy and more.
The significance of this development lies in the escalating competition in the AI arena, with Meta and Google, as well as international contenders, continually pushing OpenAI to push the boundaries of their models. Meta’s distinctive open-source approach may prove to be a key differentiator in this fierce competition.
According to reports, Meta is in the process of developing a substantial language model that will match the capabilities of GPT-4, OpenAI’s most formidable model, which serves as the foundation for ChatGPT for Business. Notably, Meta’s model will be open source and freely accessible to companies and other users.
The Wall Street Journal has indicated that this large language model (LLM) will surpass even the potency of Meta’s current flagship, Llama 2, which is also open source. This forthcoming LLM will be trained using Meta’s proprietary infrastructure, marking a departure from their previous partnership with Microsoft Azure for Llama 2 deployment. Meta has been actively constructing its data centers and procuring Nvidia H100 chips to support AI training processes.
The training of the upcoming LLM is projected to commence in 2024, with Meta aiming for a release within the same year. This development opens the door for companies, researchers, and other users to build generative AI tools using Meta’s LLMs.
Meta’s latest venture positions it squarely in competition with OpenAI, Microsoft, and Google, all of which are in a race to develop and deploy LLMs, providing generative AI services as a result.
Google DeepMind, the AI research lab of Google, is also actively working on an advanced LLM known as Gemini, which has the potential to be a strong contender in the ChatGPT space. It’s expected that Meta’s new LLM will make its debut after the introduction of Gemini.
Read the full article – AI Business
