- AI & Fries
- Posts
- Meta Unveils Llama 3.1: A Milestone in Open Source AI
Meta Unveils Llama 3.1: A Milestone in Open Source AI
+Google to provide AI training for 10,000 Indian start-ups

Fact: Our solar system has a wall. The heliopause – the region of space in which solar wind isn’t hot enough to push back the wind of particles coming from distant stars – is often considered the “boundary wall” of the Solar System and interstellar space.
What's in Today's Menu:
Meta Unveils Llama 3.1: A Milestone in Open Source AI
Google collaborates with MeitY to provide AI training for 10,000 Indian startups
Nvidia Develops China-Compatible AI Chips Amid Export Controls
OpenAI launches GPT-4o Mini
Read time : 5 minutes
Meta
Meta has introduced Llama 3.1, described as “the first frontier-level open source AI model”.
The details:
This marks a significant step in the company’s dedication to open source AI, making advanced AI technology more accessible.
Llama 3.1 includes 405B, 70B, and 8B models.The 405B model rivals the most advanced closed models while offering better cost-efficiency.
“Starting next year, we expect future Llama models to become the most advanced in the industry,” predicts Zuckerberg.
Zuckerberg compares the evolution of AI open source systems like Linux. He believes that open source AI will follow a similar trajectory.
Meta is partnering with major tech companies like Amazon, Databricks, and NVIDIA to provide development services.
source: Meta
The significance:
By offering a powerful, open source AI model, Meta makes advanced AI technology more accessible to a broader range of developers and organizations.
Thus reducing reliance on expensive, proprietary systems.
Click Here for more
Google has unveiled a series of significant initiatives aimed at boosting AI development in India, with plans to train 10,000 startups in artificial intelligence through a partnership with MeitY Startup Hub.
The details:
At the recent Google I/O Connect Bengaluru 2024 event, Google emphasized its dedication to promoting AI innovation in India, focusing on multimodal, multilingual, and mobile AI.
Google, alongside the Gemini AI models,has introduced Gemma 2 an updated version of its open models with significant performance and safety enhancements, useful for developing multilingual solutions catering to India’s diverse languages.
On the mobile front, Google has launched the Matformer framework, designed to optimize AI models for high performance and low resource consumption on mobile devices
For developers using Google Maps Platform, Google is introducing India-specific pricing, reducing costs by up to 70 percent on most APIs and offering up to 90 percent off for those building on the Open Network for Digital Commerce (ONDC).
Generated by Copilot
The significance:
This move is part of Google’s broader strategy to democratize AI and bolster the vibrant startup ecosystem in India.
Click here to know more.
Nvidia
Nvidia is creating a version of its new AI chips for the Chinese market, compatible with current U.S. export controls. This version, tentatively named the "B20," will be distributed with Inspur, a major partner in China.
The details:
Nvidia's flagship "Blackwell" chip series, unveiled in March, will begin mass production later this year.
The B200 model, within this series, is 30 times faster than its predecessor in specific tasks like chatbot responses.
U.S. export controls on cutting-edge semiconductors were tightened in 2023 to prevent advancements aiding China's military.
China accounted for 17% of Nvidia's revenue for the year ending January, down from 26% two years prior.
Nvidia's H20 chip, its most advanced for China, initially struggled but is now seeing rapid sales growth, with over 1 million units projected for this year, valued at $12 billion.
Generated by Copilot
The significance:
The introduction of the "B20" chip, alongside the H20 model, helps Nvidia strengthen its foothold in China despite U.S. sanctions. This move addresses the growing competition from Chinese tech giants like Huawei and startups such as Enflame.
Click here to know more.
OpenAI launches GPT-4o Mini
OpenAI is levelling the playing field for AI development with the launch of GPT-4o Mini, a smaller and significantly more affordable version of its powerful language model.
The details:
The high cost of using OpenAI's full-sized models has been a major hurdle for many developers, prompting some to seek more affordable alternatives such as Google's Gemini 1.5 Flash and Anthropic's Claude 3 Haiku. Addressing this issue, OpenAI is now providing a more budget-friendly option with the launch of GPT-4o Mini, making advanced AI technology more accessible.

source:OpenAI
The significance:
GPT-4o Mini, now accessible to ChatGPT users on Free, Plus, and Team plans, offers enhanced capabilities compared to GPT-3.5 while being much more cost-effective.
Click here to know more.
Tools to try
Dialogflow :A comprehensive platform developed by Google for building conversational interfaces, including chatbots and virtual assistants
Copy AI : A freestyle copywriting tool to create blog posts and more
Synthesia : Create studio-quality videos with AI avatars and voiceovers in 130+ languages.
Perplexity : A free AI-powered answer engine that provides accurate, trusted, and real-time answers to any question.
Wordtune : An AI writing assistant that helps you write high-quality content across emails, blogs, ads, and more.
Seasoned Insights
Turnitin Enhances Critical Thinking with New AI Paraphrasing Detection Feature. Read more
Scientists create AI tool that helps predict Alzheimer's with up to 82% accuracy. Read more
Meta Suspends AI Development in EU and Brazil Over Data Usage Concerns. Read more
Google, Microsoft, Infosys, IIM B and Indian AI startups form new major coalition for Responsible AI in India. Read more
AI demand displaces non-AI roles: Economic Survey 2024 highlights need for broader skill set. Read more
Remote Vacancies in Tech
ML Engineer at Desia, UK
Artificial Intelligence Engineer at Bondex, UK
ML Engineer (Generative AI Specialist) at Aimpoint Digital, UK
AI Developer at CGI, UK
Data Scientist at Kipi.bi, US
Data Analyst (Entry Level) at Engtal, US
That’s a wrap! Don't fret, there's lots more coming!
❤️ Your feedback is essential for improving our newsletter.
📣 Spread the word—share our newsletter with your friends and brighten our weekend.
🚀 Want to reach over 1000+ founders, engineers, and content creators?Advertise with us!