Meta has just unveiled a major update for its artificial intelligence assistant platform, Meta AI, built using the highly anticipated Llama 3 language model (LLM). The company claims it's "now the smartest AI assistant you can use for free." When it comes to use cases, the company showcases the ability to help users study for exams, plan dinner menus, and schedule outdoor nights. You know the drill. It's an AI chatbot.
However, Meta AI has expanded almost everywhere across the company's entire portfolio, following a beta run. It is still available on Instagram, but now users can access it on Messenger, Facebook updates, and WhatsApp. The chatbot also has a dedicated website at . You don't need a company login to use it this way, though it won't produce images. Those who recently graduated also integrate with the bot, with Quest earbud integration coming soon.
Speaking of image creation, Meta says it's now much faster and can generate images while typing. It also handles personalized animated GIFs, which is pretty cool. Hopefully, it can successfully create images of different ethnicities. We discovered this a few weeks ago, as it seemed fixated on creating images of people of the same ethnicity, even when requested otherwise.
Meta is also expanding global availability with this update, as Meta AI is now available in more than thirty countries outside the US. These include Australia, Canada, Ghana, Jamaica, Pakistan, Uganda, and others. However, there is one major caveat. It's only in English, which doesn't seem very useful for a global audience, but that might change.
Regarding safety and trustworthiness, the company claims Llama 3 underwent training on an extensive data set compared to Llama 2. It also used synthetic data to create lengthy documents to train on and claims it did not include any data sources known to contain "large amounts of personal information about individuals." Meta says it conducted a series of evaluations to see how the chatbot would handle risk areas such as discussions on weapons, cyber attacks, and child exploitation, and adapted as needed. In our brief test with the product, we've already encountered discrepancies, as shown below.
Artificial intelligence has become one of Meta's CEOs, along with secluded Hawaiian cattle ranch, but the company is still playing with OpenAI and to a lesser extent, Google. Meta's Llama 2 never really excited users, due to its limited feature set, so maybe this new version of the AI assistant will hit the mark. At the very least, it should be able to paint the town red, or more accurately, botch someone else's painting of painting the town red.
