By Natalia Rueda 
As artificial intelligence continues to expand its reach, Microsoft and Meta have made a bold move to enter the industry with their new open source AI model, Llama 2. The new model promises to be on par with GPT-4 and competitors OpenAI and Google, who have both developed free versions of their AI language models.
Llama 2 is the first free AI that is available to researchers and companies. It was trained on forty percent more data than Meta’s first model, Llama, with two billion tokens and a million human annotations. It is able to run in either local or cloud-based mode on Windows, Amazon Web Services (AWS) and Hugging Face.
It is unsurprising that Microsoft is involved with Meta, as there has been an investment of more than $10 billion into OpenAI, Microsoft’s preferred partner. However, these two tech giants have paired up to become a preferred customer of Meta.
With this move, Meta is aiming to compete with OpenAI and Google, taking back some of the ground lost in the AI race. They hope to do this by utilizing the model’s communal stress testing, so that potential and existing problems can be identified and solved quickly.
But this is still only the beginning for Meta. It needs to prove itself to be a genuine competitor and this new input of AI could be the ticket. It offers the potential opportunity to develop a series of high-level applications akin to ChatGPT from OpenAI and Microsoft Bing’s AI.
Only time will tell how successfulMeta’s Llama 2 will be in this game of AI. But, with Microsoft’s backing and Meta’s commitment to open source, it stands a good chance of raising the stakes in this new wave of AI development.