Meta to launch its own AI system soon to take on rival Open AI's GPT

Formerly known as Facebook, Meta is preparing to develop a new AI system that it thinks will compete in computational power and complexity with OpenAI's GPT systems. The system will extend Meta's current Llama models and bring Meta up to speed in the generative AI competition. The anticipated system, the details of which are still being worked out, would assist other companies in developing services that generate complex text, analysis, and other output.

According to a report, Meta's new AI model will compete with the capabilities of OpenAI's ChatGPT-4 chatbot. It is said that Meta?is substantially investing in AI capabilities and growing its infrastructure. According to the Wall Street Journal, training for this new Large Language Model (LLM) would most likely begin in early 2024. Furthermore, Meta has assembled a dedicated team to work on this AI model, with a focus on creating AI tools that can accurately replicate human expressions.

Mark Zuckerberg, Meta's CEO, is expected to make the new AI Model available to enterprises and researchers for free at first. This method appears to be congruent with Meta's goal of ‘democratising AI’ making it freely available for a variety of uses.

What is Meta’s Llama?

Earlier, Meta created Code Llama, an AI tool for creating new code and debugging human-written code. Code Llama is a large language model (LLM) that uses text prompts to generate and discuss code. Its purpose is to simplify developer operations and make learning more accessible to aspiring coders.

Designed to produce and analyse code snippets, Code Llama assists with finding solutions. There are?three versions of the Code Llama base, each containing 7 billion, 13 billion, and 34 billion parameters. The tool?works similarly to Google for code snippets, producing full, active codesets in response to text requests.

This saves?a significant amount of time. While code expertise is essential for debugging, as previously said, most programmers still look for code examples for specific aspects, and?then add them to the mix, albeit in customised format.

According to reports, Meta is now collecting more Nvidia H100 chips to power its AI goals. Nvidia processors are currently in high demand and limited supply, which means that the business with the most chips has the best chance of winning the broader AI development race.

?Users’ Expectation: What’s new unfolding?

It will be interesting to see how Meta intends to deploy its own AI and LLM technologies, as well as how they will be incorporated into its in-app experiences. Meta is already experimenting with artificial intelligence-powered aspects such as graphic creation tools and an Instagram chatbot.

Perhaps, with a more powerful AI system, it will try to augment and optimise these experiences, while also incorporating higher-end AI into its ad targeting and generating process.

Overall, it's not surprising to see Meta making a greater play in the AI market, considering its long-term work in the field and the potential it has to compete based on its treasure of conversational data and interaction. Stay tuned to CogentIBS newsletters for more such tech updates!