It will be present almost everywhere.
A significant update has recently been revealed by Meta for their artificial intelligence assistant platform, which is known as Meta AI. This update was developed with the long-awaited open source Llama 3 large language model (LLM). It is now “the most intelligent artificial intelligence assistant you can use for free,” according to the business. From the perspective of use case scenarios, the company emphasizes its capacity to assist users in preparing for examinations, organizing dinners, and organizing nights out. It is not a mystery to you. It is an artificial intelligence chatbot.
Meta AI, on the other hand, has since grown into virtually every nook and cranny across the entirety of the company’s portfolio, following a trial run with direct messages on Instagram the previous week. It is still accessible on Instagram, but users may now access it through Messenger, Facebook feeds, and Whatsapp as well. There is also a dedicated web portal for the chatbot, which can be found at, wait for it, meta.ai. It is not necessary to have a company login in order to utilize it in this manner; nevertheless, it will not produce photos. The recently announced Ray-Ban smart glasses are also compatible with the bot, and the Quest headset will soon be able to take use of this connectivity.
Meta claims that it is now significantly faster and will make images as you type. This is in reference to the topic of image generation. A interesting feature is that it can also handle animated GIFs that have been customized. That it is able to successfully generate images of people of diverse races is something that we hope for. A couple of weeks ago, we discovered that it had difficulty with this fundamental idea, since it appeared to be biased toward making photos of people of the same race, even when it was directed to create images of people of different races.
Additionally, in conjunction with this update, Meta is expanding its global availability, as Meta AI will soon be available in more than a dozen countries outside of the United States. The countries of Australia, Canada, Ghana, Jamaica, Pakistan, Uganda, and others are included in this category. Having said that, there is a significant warning. Whatever the case may be, it is only available in English, which may not seem particularly relevant to a global audience.
When compared to Llama 2, the business claims that Llama 3 has been trained on a larger data set, which is a significant improvement in terms of both safety and reliability. Additionally, it made use of synthetic data in order to generate lengthy papers for the purpose of training, and it asserts that it excluded any data sources that are known to include a “high volume of personal information about private individuals.” Meta claims that it carried out a series of studies to determine how the chatbot would respond to potentially dangerous situations, such as chats concerning firearms, cyber attacks, and the exploitation of children, and that it made adjustments as necessary. As may be seen in the following image, we have already experienced hallucinations while using the product for a short period of time.
Mark Zuckerberg, CEO of Meta, has made artificial intelligence (AI) one of his pet projects, along with breeding extremely costly cattle for meat on a remote Hawaiian estate. However, the business is still playing catch-up to OpenAI and, to a lesser extent, Google. Because of its limited feature set, Meta’s Llama 2 never truly impressed consumers. However, it is possible that this next version of the artificial intelligence assistant will be able to capture lightning in a bottle. The absolute least that it should be able to do is draw lightning in a bottle, or more correctly, it should be able to slightly modify the drawing of lightning in a bottle that was done by someone else.