Alibaba’s Cloud AI Ambition Soaring with Meta’s Llama2

·

·

Alibaba and The Cloud Computing Crusade

Let’s address Alibaba’s cloud computing division. The proud first Chinese enterprise to ride on the coattails of Meta’s open-source artificial intelligence model Llama. Giving the reins of developing programs to its Chinese business users, it’s the very picture of generous.

Now let’s not overlook the other Llama in the pen. Meta recently let loose Llama2, the commercial version that’s being sold as the potent yet free alternative to OpenAI and Google’s pricey proprietary models. And who, you ask, was Meta’s favorite kid in the playground for Llama2? Microsoft, of course, but Alibaba isn’t far behind, having hitched its wagon to this new star.

Let’s be honest, Alibaba has played a strategic move here. Launching the first training and deployment solution for Llama2 in China – a call to arms for all developers to flex their creative muscles on Alibaba Cloud. For a firm eyeing a stock market listing amidst escalating competition, this could be just the ticket. Plus, with Meta in their corner, they could potentially have a roster of committed clients for their cloud business.

The Crossfire of AI Ambition

Our story takes an intriguing turn. The US has been pulling all stops to restrict Chinese access to many homegrown AI technologies, especially those pesky AI semiconductors. Enter Llama2, Alibaba’s ticket to staying on the AI frontlines.

Meta, on the other hand, sees a silver lining in China, the world’s second-largest economy, where its social media platform has been persona non grata. But no comments from either camp – Meta has zipped its lips, and Alibaba Cloud seems to have lost its voice.

The AI Arms Race

China is no stranger to ambition. Eager to give Uncle Sam a run for his money in the AI field, Beijing has been urging Chinese companies to fast-track the development of their own ‘controllable’ AI models. Riding the wave are Alibaba and Tencent Holdings, who have been fueling their AI engines in recent months.

Llama2, akin to the models behind popular chatbots like OpenAI’s ChatGPT and Google’s Bard, is a trained machine learning model fed on vast data. With the ability to churn out coherent and natural-sounding outputs, Llama2 is free for companies with less than 700 million monthly active users. Those above will have to get on their knees to Meta for a license.

With a slight caveat, if clients plan to use Llama2 to serve the Chinese public, they’ll need to adhere to Chinese laws and steer clear of content that might rub the country the wrong way.

Source: www.reuters.com