Meta’s Llama 2 is no short of a success for open source. Now, the company is ready to launch the next version of it as soon as the first quarter of 2024.
According to several speculations, Meta will launch Llama 3 early next year. The best part about it is that it is going to be open source for research and commercial use as well. Meta has also highlighted that it should be deployed responsibly, and the company will frame policies and mechanisms to use it ethically and responsibly.
Moreover, according to the same leaks, Meta has also partnered with Dell to offer Llama 2 on-premises for enterprise users for ensuring more control and security over personal data.
It seems probable since OpenAI at the DevDay has announced GPT-4 Turbo and Google is also planning to launch Gemini. It is about time that Meta actually launched Llama 3, since it is already in training.
Interestingly, Mark Zuckerberg, in his latest podcast with Lex Fridman in the metaverse, said that Meta might have to reconsider if it is going to open source the next iteration of Llama, which is Llama 3. “Right now, the priority is building that into a bunch of consumer products,” said Zuckerberg.
But now, it seems like Meta wants to go the open source way again.
In the podcast, Zuckerberg added that Meta trained Llama 2 and released it as an open source, but it is not a consumer product, but just an AI infrastructure. Though Zuckerberg is all about open sourcing AI and loves what the community has been doing with Llama 2, when it comes to Llama 3, he said that the debate with Fridman was very helpful for open sourcing Llama 2, and the same would be needed for Llama 3.
“We would need a process to red team this, and make it safe. My hope is that we would be able to open source the next version when it is ready to, but we are not close to doing that this month. It’s a thing that we are still early in work now,” said Zuckerberg. This time, Llama 3 might even be better than GPT-4.
Given all the debates about AI policies and the recent Biden order on regulating AI, a user on Reddit says, “All they need to do is make it 180B and most people will have no way to abuse it.”