When Meta initially hinted in April that it was developing an open-source model with performance comparable to the top private models from firms like OpenAI, it was a first for the AI business.
That model is here today. The largest-ever open-source AI model, Llama 3.1, is being released by Meta. According to the company, it performs better on multiple benchmarks than Anthropic’s Claude 3.5 Sonnet and GPT-4o. Along with expanding the Llama-based Meta AI assistant’s language and country of availability, it also adds a capability that allows it to create graphics based on a person’s unique likeness. By the end of this year, CEO Mark Zuckerberg now believes Meta AI will overtake ChatGPT as the most popular assistant.
Also Read: How to use Imagine Me? Meta AI Tool for Selfie (Text to Image) Generation
Compared to the smaller Llama 3 models that were released a few months ago, the Llama 3.1 is far more complicated. With 405 billion parameters, the largest version was trained using more than 16,000 of Nvidia’s incredibly expensive H100 GPUs. Although Meta has not disclosed the exact cost of building Llama 3.1, it is reasonable to assume that it was in the hundreds of millions of dollars, just based on the price of the Nvidia chips.
So why is Meta still giving Llama a free license that just needs the consent of businesses with hundreds of millions of users, considering the expense? Similar to how Linux evolved into the open-source operating system that now powers the majority of phones, servers, and devices, Zuckerberg contends in a letter posted on Meta’s corporate blog that open-source AI models will surpass proprietary models and are currently advancing more quickly.
He draws a parallel between Meta’s previous Open Compute Project, which he claims saved the company “billions” by enlisting the assistance of outside firms like HP to enhance and standardize Meta’s data center designs as it was expanding its capacity, and its current investment in open-source AI. He writes, “I believe the Llama 3.1 release will be an inflection point in the industry where most developers begin to primarily use open source.” He anticipates the same scenario to play out with AI in the future.
Also Read: How to use Meta AI in WhatsApp, Instagram, Facebook, and Messenger to Get Rapid Answers
Meta is collaborating with over twenty organizations, such as Microsoft, Amazon, Google, Nvidia, and Databricks, to assist developers in releasing their versions of Llama 3.1 into the wild. According to Meta, the manufacturing running costs of Llama 3.1 are around half of those of OpenAI’s GPT-4o. The model weights are being made available so businesses may train and fine-tune them using their data.
Also Read: Meta Unveils JASCO: Advanced Text-to-Music Model with Fine-Grained Control and High-Quality Output
It should come as no surprise that Meta is keeping quiet about the data it used to train Llama 3.1. AI industry employees claim that this information is kept confidential as trade secrets, but detractors accuse them of using this as a ploy to postpone the impending wave of copyright litigation.
Meta will state that the 405-billion parameter version of Llama 3.1 outperformed the smaller 70-billion and 8-billion versions through the use of synthetic data, or data produced by a model instead of by humans. Llama 3.1 will be well-liked by developers as “a teacher for smaller models that are then deployed” in a “more cost-effective way,” according to Meta’s Vice President of Generative AI, Ahmad Al-Dahle.
Red teaming, also known as adversarial testing, of Llama 3.1 by Meta included searching for possible applications in biochemistry and cybersecurity for the first time. The “agentic” behaviours that Meta is portraying as arising are another incentive to test the model more rigorously.
Also Read: 10 Best Metaverse Virtual Worlds to Watch in 2024
Meta’s AI assistant, Llama, is positioned as a general-purpose chatbot similar to ChatGPT and is integrated into nearly every aspect of Facebook, WhatsApp, Instagram, and other platforms. Llama 3.1 will be available in the US starting this week on WhatsApp and the Meta AI website. In the upcoming weeks, Instagram and Facebook will also be able to access them. Additional languages like French, German, Hindi, Italian, and Spanish are also being supported by updates.
Although the most sophisticated 405-billion parameter model of Llama 3.1 is available for free usage in Meta AI, the assistant will automatically switch to the more basic 70-billion model once you exceed a certain number of prompts in a given week. This implies that Meta cannot operate at full capacity due to the 405-billion model’s high cost.
Meta AI’s new “Imagine Me” function allows you to add your likeness to pictures it creates by having it scan your face using the camera on your phone. With this method of capturing your likeness rather than using the images on your profile, Meta is hoping to prevent the development of a deepfake machine. The business observes a desire among consumers to produce more AI-generated content and distribute it to their feeds, even if doing so obfuscates the distinction between what is and is not real.
Also Read: Meta 3D Gen: New Algorithm Converts Text Descriptions into 3D Models
With over 100 million users on ChatGPT, Zuckerberg predicted that Meta AI would surpass all other chatbots by the end of the year. However, Meta has not released any usage statistics for its helper. Al-Dahle asserts, “I think the industry as a whole is still early on its path towards product market fit.” Meta and other players believe the race is just getting started, despite how overhyped AI might already feel.
