It’s turning into a significant week in the AI arms race, with Meta launching a web chatbot that’s similar to ChatGPT, expanding its AI within Facebook, Messenger, Instagram, and WhatsApp, and rolling out a new model, Llama 3.
Almost four billion people – roughly half the world’s population – use one of Meta’s products. However, not everyone will get access to the changes just yet.
In this guide, we’ll explain how you can access Meta AI via the company’s social media platforms, take a peak at the web app, and discuss a significant blunder already made by Meta AI since it became more widely available.
How to Use Meta AI on Facebook, WhatsApp and Instagram
You can now use Meta AI by simply searching for anything while on Facebook Messenger, Facebook, Instagram, and WhatsApp. So, instead of having to leave the app to Google something, you can simply “pop up” to Meta AI like you would a friend.
However, you can also “@” meta AI when typing in a group chat Message to ask it for recommendations, information, and ideas.
It’s not available everywhere just yet. Along with the US, users in Australia, Canada, Ghana, Jamaica, Malawi, New Zealand, Nigeria, Pakistan, Singapore, South Africa, Uganda, Zambia, and Zimbabwe will be the first to get their hands on it.
Excitingly, Meta says it also plans to “make image generation faster”, so is letting users “create images from text in real-time using Meta AI’s Imagine feature”.
🔎 Want to browse the web privately? 🌎 Or appear as if you're in another country?
Get a huge 86% off Surfshark with this special tech.co offer.
This feature is being rolled out as a beta for WhatsApp and is incorporated into the new web tool – however, you will need to log in to generate imagery, as well as seeing your conversation history.
Along with being integrated into Meta’s social media platforms, for the first time, you can now use Meta’s AI chatbot as a standalone tool, just like you can use ChatGPT.
How to Use Meta’s Chatbot, With Llama 3
Meta AI launched this week, accompanied by CEO Mark Zuckerberg’s bold claim that it’s “the most intelligent AI assistant that you can freely use.” It’s now powered by Llama 3, Meta’s most intelligent, publicly available Large Language Model (LLM).
The interface, response formatting, and color scheme make it look like a straight cross between ChatGPT and Google’s Gemini:
I asked Meta AI to plan a trip to Athens for me – and I think it did pretty well! Image: Tech.co
Although Meta will encourage you to log into Facebook when you type your first response into the chatbot, it’s possible to use it without logging in at the moment too. However, it will prompt you again every time you start a new conversation.
Unlike the social media integrations, Meta AI is available to everyone – all you have to do is navigate to the page via a search engine.
Meta’s AI Already Makes Blunder
AI tools have tended to cause controversy right after they’re launched – Bard (now Gemini) made a factual error during its first public demo, while X’s chatbot Grok drew the ire of the platform’s right-wing commentators for its “woke” responses.
Meta AI seems like it’s going to be no different – there are already reports of strange happenings taking place on Facebook.
Sky News reports that Meta AI was spotted claiming to have raised a disabled child in response to a post on a private parenting group. However, the chatbot apologized and corrected itself shortly after being questioned by members.
It’s unlikely this is the last weird answer we’ll see from Meta AI – especially considering how deeply integrated into the world’s biggest social media sites the technology now is.
