Meta Releases Llama 2 for Free and Apple's Testing Apple GPT

All aboard!

We're welcoming our newest member to the AI hype train: Apple!

But before we get there and share what they're doing with generative AI, our next stop is Meta.


Meta's giving away its AI tech for free to try to beat ChatGPT and announced it's open-sourcing its large language model, Llama 2.

Yes, you read that right.

Zuck, in a partnership with Microsoft, made Llama 2 free to use (even commercially) so that businesses, startups, and researchers can access it and develop their own AI tools as a community.

Meta said in its press release that Llama 2 was trained on 40% more data compared to version 1.0. It "outperforms" other large language models (LLMs) when it comes to logic reasoning, coding, proficiency, and knowledge tests.

Llama 2 is available on Microsoft's Azure platform, and you can run it locally on Windows. But that's not all!

You can access it across different platforms like Amazon Web Services, Hugging Face, and other providers.

The main takeaway here is that Meta and Microsft have made generative AI technology open to everybody. Who knows what people will develop now that it's open to the public.

Now Apple, on the other hand, is testing an AI chatbot but has no idea what to do with it.

Engineers at the company are calling it 'Apple GPT,' and it uses its own LLM called Ajax, which runs on the Google Cloud and is built with Google JAX.

Apple has multiple teams working on it, using the chatbot internally and testing it for potential privacy implications.

But compared to the other tech giants in the generative AI space, Apple has been super quiet.

Apple's Siri may have a little bit of AI tech in it, but the voice assistant still has a long way to go.

Tim Cook, CEO of Apple, said the company is "looking at [AI tech] closely" but also expressed concerns that some issues need to be sorted out.

So we don't know what Apple has in store yet, but they plan on making a "significant AI-related announcement" next year.

We'll just have to see if the AI hype train will still be going strong or derail by then.