Facebook, Instagram and WhatsApp owner Meta Platforms is making a massive push into the generative AI space.
It’s now rolling out its free Meta AI in English virtual assistant in more than a dozen countries outside of the US including Australia, Canada, Ghana, Jamaica, Malawi, New Zealand, Nigeria, Pakistan, Singapore, South Africa, Uganda, Zambia and Zimbabwe.
This week, it released the early versions of its latest large language model, Llama 3, and an image generator that updates pictures in real time while users type prompts. The latest models will be integrated into its virtual assistant Meta AI.
Meta AI is being rolled out in search functions across Facebook, Instagram, WhatsApp and Messenger. It will access real-time information from across the web without having to switch between apps.
Meta AI access is also available when you’re scrolling through your Facebook Feed. If you come across a post you find interesting, you can ask Meta AI for more info from the post itself.
An image generation feature is being rolled out first in the US to begin with. Users there can create images from text in real-time using Meta AI’s Imagine feature.
Apart from being integrated into those platforms, Meta AI also has a new dedicated website that is available in Australia too.
The landing page of the new website invites users to help the AI assistant create a vacation packing list, help make an email sound more professional, play 1990s music trivia with them, or paint pictures of the New York City skyline.
These are concerted efforts to have Meta compete more aggressively with Microsoft’s OpenAI ChatGPT which has grown exponentially. Meta has been widely releasing its Llama models for use by developers building AI apps.
Meta has equipped Llama 3 with new computer coding capabilities and fed it images as well as text while training it, although the model will output only text. More advanced reasoning abilities are expected to follow in subsequent versions of Llama 3, reports Reuters. Future releases will also include “multimodality,” meaning they can generate both text and images.
Meta notes that Llama 3 models will soon be available on AWS, Databricks, Google Cloud, Hugging Face, Kaggle, IBM WatsonX, Microsoft Azure, NVIDIA NIM, and Snowflake, and with support from hardware platforms offered by AMD, AWS, Dell, Intel, NVIDIA, and Qualcomm.
It seems aware of the safety concerns around the rapid rolling out of this powerful too. To that end, it has introduced new trust and safety tools with Llama Guard 2, Code Shield, and CyberSec Eval 2.
