Meta has taken a big step in bringing powerful AI to our phones. They've made smaller, faster versions of their Llama AI models. These models run great on smartphones and tablets, opening up new ways to use AI.

        The new models are super fast and use less memory. They were made using a technique called quantization, which makes the math behind AI simpler.

        Testing on Android phones showed these models were much smaller and used less memory. They handled texts up to 8, 000 characters, which is plenty for most mobile apps.

        This opens up a whole new world. Now, we might see AI doing tasks right on our phones, like summarizing documents or helping with creative writing. This could make our phones smarter and keep our data private.

        But this isn't just about making our phones smarter. It's also about the tech race. Companies like Google and Apple are careful about how AI runs on their devices. Meta is taking a different approach by working with chip makers and making the AI models open-source. This could help developers create new apps quickly.

        The future of AI on our phones looks exciting. But it's not guaranteed. These models need powerful phones to work well. And there are other companies with their own plans for AI on phones. Still, it's clear that AI is moving away from data centers and into our pockets.

Meta's Move: AI on Phones Becomes Reality

Actions