Meta's new language model fuels Big Tech's AI arms race.

[{"selector":"#anim-ae15aac9-4fa0-48f4-8b38-6cddedaf0b0c","keyframes":{"opacity":[0,1]},"delay":0,"duration":2000,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}]

What's LLaMA and how does it work?

NEWS With the release of Microsoft-backed OpenAI's ChatGPT late last year, the public competition to rule the AI technology arena has began.

[{"selector":"#anim-42b40898-a712-42e8-98f5-fa404cb220c2","keyframes":{"opacity":[0,1]},"delay":0,"duration":2000,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-eab79ce3-db34-4831-974e-4c877b9d4a75 [data-leaf-element=\"true\"]","keyframes":{"transform":["translate3d(7.962100808745443%, 0, 0)","translate3d(0%, 0, 0)"]},"delay":0,"duration":2000,"easing":"cubic-bezier(.3,0,.55,1)","fill":"both"}]

NEWS This drove tech giants like Alphabet Inc. (GOOGL.O) and China's Baidu Ltd. (9888.HK) to promote their own products.

[{"selector":"#anim-cb2a8e12-017d-445a-9b53-6b1ecefcc86b","keyframes":{"opacity":[0,1]},"delay":0,"duration":2000,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-c518b2e8-087f-426b-9293-9f1303a7f7c5 [data-leaf-element=\"true\"]","keyframes":{"transform":["translate(7.301065662149618%, 0%) scale(1.5)","translate(0%, 0%) scale(1)"]},"delay":0,"duration":2000,"easing":"cubic-bezier(.3,0,.55,1)","fill":"forwards"}]

NEWS Meta's LLaMA, otherwise known as Large Language Model Meta AI is an new language introduced by Facebook.

[{"selector":"#anim-c1a2ce16-6596-42fd-aff1-42c913d87962","keyframes":{"opacity":[0,1]},"delay":0,"duration":2000,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-06f15ace-6771-4731-a4d6-3f832ae54ec5 [data-leaf-element=\"true\"]","keyframes":{"transform":["translate(0%, 0%) scale(1.5)","translate(0%, 0%) scale(1)"]},"delay":0,"duration":2000,"easing":"cubic-bezier(.3,0,.55,1)","fill":"forwards"}]

NEWS Researchers and organizations connected to the government, civil society, and academia will be able to use Meta's LLaMA.

[{"selector":"#anim-47714b1a-29ec-4338-a276-94106aeef874","keyframes":{"opacity":[0,1]},"delay":0,"duration":2000,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-5082057f-3756-4356-9dd3-ca8f66f244f0 [data-leaf-element=\"true\"]","keyframes":{"transform":["translate(8.199708643365838%, 0%) scale(1.5)","translate(0%, 0%) scale(1)"]},"delay":0,"duration":2000,"easing":"cubic-bezier(.3,0,.55,1)","fill":"forwards"}]

NEWS To summarize data and provide content, large language models mine enormous quantities of text. BUT!

[{"selector":"#anim-5b553192-e0ac-484c-b33f-a5d8a4cf8b22","keyframes":{"opacity":[0,1]},"delay":0,"duration":2000,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-baa51233-915b-4a7e-898d-447b4ef79150 [data-leaf-element=\"true\"]","keyframes":{"transform":["translate(0%, 0%) scale(1.5)","translate(0%, 0%) scale(1)"]},"delay":0,"duration":2000,"easing":"cubic-bezier(.3,0,.55,1)","fill":"forwards"}]

NEWS Meta's LLaMA is trained on 20 languages. Meta said, and uses "much less" computational resources than earlier solutions.

[{"selector":"#anim-55f6876a-0da4-46f0-aab7-8a2a601aae92 [data-leaf-element=\"true\"]","keyframes":{"transform":["translate(0%, -0.000003569388940860563%) scale(1.5)","translate(0%, 0%) scale(1)"]},"delay":0,"duration":2000,"easing":"cubic-bezier(.3,0,.55,1)","fill":"forwards"}] [{"selector":"#anim-7af750ea-be4c-4125-8835-7c643d54b367","keyframes":{"opacity":[0,1]},"delay":0,"duration":2000,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}]

NEWS Senior software analyst at D.A. Davidson, said Meta's statement "seems to be a step in proving their generative AI capabilities to apply them into their products in the future."

[{"selector":"#anim-59e59611-8d7e-4c41-a995-a721174f5fa6 [data-leaf-element=\"true\"]","keyframes":{"transform":["translate(0%, 0%) scale(1.5)","translate(0%, 0%) scale(1)"]},"delay":0,"duration":2000,"easing":"cubic-bezier(.3,0,.55,1)","fill":"forwards"}] [{"selector":"#anim-233128ed-d845-4c63-bd80-6e6d7c329189","keyframes":{"opacity":[0,1]},"delay":0,"duration":2000,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}]