They’ve done it again.
Finally, Meta has released LLaMa 3, a new generation of large-scale language models (LLM) for Meta Platforms. And they did not disappoint (well, except for one thing).
But apart from their overall impressive accomplishments, one discovery they made will likely eclipse the others. While it’s certainly a discovery that really made Mark Zuckerberg happy, they’re not the only big tech companies to benefit from it.
You’re probably tired of AI newsletters that tell you how this or that happened **only**. And there are a lot of these newsletters because it’s easy to talk broadly about events and things that have already happened.
But newsletters that tell you what’s coming are rare. Are you interested in easy-to-understand insights that look ahead to the future of AI? The Tech Oasis A newsletter might be perfect for you.
🏝️🏝️ Subscribe now below:
Those who like the open source movement will love the LLaMa 3 model.
something that fills the gap
Meta has released two models.
- Llama 3 8Ban 8 billion parameter model with a knowledge deadline of March 2023 (i.e., if you ask about events after that deadline, you won’t know anything)
- Rama 3 70Ba 70 billion parameter model with a knowledge cutoff of December 2023.
But now, Meta is going one step further and officially bringing these models to life. They are releasing Meta AI. This is a free virtual assistant that you can interact with through Meta’s most popular products. whatsapp, Facebook Messengeror Instagram.
In other words, this time they are confident that the model is good enough to offer to billions of users, and this shows how much they trust this new model. …
