Interpretation of the digital asset policy of the Google Store: NFTs are allowed in apps and games, and mining on devices is not allowed

Author: Zhao Ying

The AI large-scale model competition is heating up, and Meta came to disrupt the situation and commercialize the open-source large-scale model.

On Tuesday, according to the Financial Times, **Meta is preparing to release a commercial version of its large language model LLaMA, which will allow startups and enterprises to build custom AI software on top of the technology. **

The move would mean pitting Meta head-to-head with the likes of Microsoft-backed OpenAI and Google, which are far ahead in the race to develop generative artificial intelligence.

Meta VP and chief AI scientist Yann LeCun said at a conference in Aix-en-Provence on Saturday:

The AI competitive landscape will completely change in the coming months, maybe weeks, when open source platforms will emerge that are as good as non-open source platforms.

The open source LLaMA model is about to be commercialized

The previous article pointed out that in February this year, Meta released its open source large language model LLaMA, which can help researchers develop proprietary AI software at low cost, thus causing an uproar in the AI field. The new version will be made more widely available and customizable by businesses, three people familiar with the matter said.

In stark contrast to competitors such as OpenAI, LLaMA is an open-source model. Essentially, Meta presents its AI technology as open-source software (computer code that can be freely copied, modified, and reused), which gives outsiders everything they need to quickly build their own chatbots.

**The commercialization of LLaMA is not only beneficial to its own competition. AI developers hope to find products that can replace the proprietary software sold by competitors Google and OpenAI. It will also indirectly benefit Meta's AI development. **

The benefits of an open source model include a higher acceptance rate from users, which then feeds in more data for the AI to process. The more data an LLM has, the more powerful it becomes. Additionally, the open-source model enables researchers and developers to discover and address vulnerabilities while improving technology and security, and tech companies like Meta have faced scrutiny over the years for various privacy and misinformation scandals.

While Meta's technology is open source and currently free, the company has been exploring commercialization, two people familiar with the matter said. There are currently no plans to charge, and Meta will not do so in an upcoming release, a person familiar with the matter said.

Joelle Pineau, Meta's vice president of AI research, declined to comment on commercialization, but said:

This stuff is indeed open source, but you don't completely give away intellectual property rights to the work.

The competitive landscape of large models may undergo drastic changes

While giving software away for free might seem antithetical to making money, experts believe businesses can also use the tactic to open up markets. A person familiar with Meta's thinking said:

**Meta has realized that they are behind in the current AI hype cycle, but this gives them a way to open up the ecosystem and they seem to be doing the right thing. **

Earlier this week, Meta's head of global affairs, Nick Clegg, extolled the benefits of open source, saying:

Open source is the best antidote to AI-related fears. Of course, being open source also helps Meta catch up to competitors, as it allows companies of all sizes to refine the technology and build applications on top of it.

Meta would also be at risk, though, and could face legal action if it is found to have used copyrighted material to train the software in the process of commercializing the next version of LLaMA.