Meta set to enter AI space with a ChatGPT-4 rival

Meta is reportedly working on a new AI model that can outperform OpenAI's ChatGPT-4 and other popular AI bots including Microsoft Bing and Google Bard. So, it is safe to say that the Mark Zuckerberg-led tech giant is finally on the verge of jumping in on the AI chatbot bandwagon.

Likewise, a slew of other big tech companies are sparing no effort to take advantage of the emerging tech. Despite ChatGPT traffic falling for the third month in a row, it is safe to say OpenAI is currently at the forefront of the AI segment.

Google and Microsoft aren't lagging behind as well. Apple and Meta, on the other hand, are still mum on their plans to make their foray into the booming segment. According to earlier reports, Apple is investing a considerable amount of money in the AI space every day.

Moreover, the word on the street is that Apple is working on multiple AI projects of its own. While the Cupertino-based tech giant has neither confirmed nor denied this speculation yet, a new report by The Wall Street Journal claims Meta is developing a more powerful AI system than OpenAI's most advanced model.

Meta set to follow in OpenAI's steps

Earlier this year, a report indicated that OpenAI is working on an advanced AI model that will use 10 million NVIDIA GPUs to draw its power. Similarly, the WSJ report claims Meta is buying several Nvidia H100 AI-training chips to build an infrastructure that will support its power-packed AI model.

According to the report, Meta's upcoming AI system will be capable of giving OpenAI's most advanced LLM (GPT-4) a run for its money. On top of that, Meta is also ensuring that it doesn't have to rely on Microsoft in terms of training its AI system. Reportedly, a group of people is working on AI tools that copy human expressions.

What is ChatGPT-4?

If the recently surfaced report is anything to go by, Meta will begin training its upcoming AI model in early 2024. The report claims Meta is hoping that its upcoming LLM (large language model) will be more powerful than the currently dominating GPT-4 from OpenAI.

It is worth noting that ChatGPT's premium version uses GPT-4, making it the most sophisticated generative AI chatbot that can perform several tasks without breaking a sweat. For instance, the AI bot can tackle complex problems, compose poetry and even write code.

To recap, ChatGPT-4 was introduced back in March. According to OpenAI, this large multimodal model can accept image and text inputs and emit text outputs. While the LLM is less capable than humans in many real-world scenarios, it showcases human-level performance on several professional and academic benchmarks.