Meta Unleashes New Powerful Free AI Model Llama 3.1
Hi there.
This week, Meta’s new AI model Llama 3.1 is making waves as the most powerful yet free model on the internet. Chat-GPT is in the news again, with a new plan to build its own AI chips, while in China, a company is using AI for the reanimation of deceased loved ones.
Also, check out my new YouTube video on how you can perform your own half-yearly review and make the best plan today to achieve your goals by December. Let’s dive in.
📢 Shoutout
Spotlight Studios: If you’re struggling to launch your podcast, my team will help you create stunning episodes and short-form content for social media.
I’ve been building my content creation team for a couple of years, and we’re ready to help new creators and businesses produce powerful content. If you’re interested, just reply to this email and I’ll schedule a call with you.
🖥️ Tech Bytes
Meta Releases Its Biggest & Best "Open" AI Model Yet
Meta has released Llama 3.1, the largest open-source AI model, outperforming competitors like GPT-4o and Claude 3.5 Sonnet. CEO Mark Zuckerberg predicts Meta’s AI assistant, powered by Llama 3.1, will surpass ChatGPT by year-end. Llama 3.1 is a significant investment with 405 billion parameters and trained using 16,000 Nvidia H100 GPUs.
Meta's strategy includes partnerships with Microsoft, Amazon, and Google for Llama 3.1 deployment. The model is cheaper to run than OpenAI’s GPT-4o. Meta is releasing model weights for customization, though training data specifics are undisclosed. Llama 3.1 uses synthetic data and can integrate with search engines and execute Python code.
Meta's AI assistant is being introduced in WhatsApp, Instagram, and Facebook, initially with the 405-billion parameter model, then a 70-billion model after extensive use. It will support multiple languages and features like AI image generation. Meta aims to integrate its AI across platforms, including its Quest headset and Ray-Ban glasses.
Despite advancements, Meta notes the AI industry is still in the early stages and evolving.
OpenAI Wants to Develop Its Own AI Chips Now
OpenAI is considering making its own AI chips due to the shortage and high costs of AI GPU chips. They are exploring options like acquiring a chipmaker or partnering with existing ones like Nvidia, which holds over 80% of the market.
CEO Sam Altman has highlighted the high costs and scarcity of these chips. OpenAI relies on a Microsoft supercomputer with 10,000 Nvidia GPUs, making each ChatGPT query cost about 4 cents. Scaling up to a fraction of Google's search volume could cost around $48.1 billion in GPUs and $16 billion annually.
Custom AI chips would align OpenAI with Google and Amazon but would need significant investment and time. Microsoft, OpenAI's main backer, is also developing a custom AI chip, which OpenAI is testing and is expected to be revealed soon.
Chinese Tech Offers AI Reanimation of Deceased Loved Ones
Chinese tech companies now offer services to create AI avatars of deceased loved ones, using advanced AI for lifelike digital representations that can speak and interact. Silicon Intelligence developed an avatar for Sun Kai, a tech executive who talks to an AI version of his late mother for emotional support.
This "resurrection" service allows clients to maintain a form of relationship with lost loved ones. The AI avatars use vast language data and advanced chatbots like Baidu’s Ernie or OpenAI’s ChatGPT, making them affordable and realistic.
Digitally cloning the deceased raises ethical questions about personal data use and psychological impacts. Companies like Microsoft and OpenAI have internal committees to address these issues, but there is no central regulatory body in the U.S. or China. The rise of "deadbots" suggests a future with more AI simulations of deceased individuals, highlighting the need for ethical guidelines and regulations.
Elon Musk Unveils World's Most Powerful AI Training Cluster
Elon Musk and his AI company, xAI, have announced the launch of the "Memphis Supercluster," which Musk claims to be the most powerful AI training cluster in the world. Utilizing 100,000 Nvidia H100 GPUs, the supercluster aims to develop Grok 3.0, slated for release in December 2024, which is expected to be the most powerful AI by every metric.
Musk also revealed that Grok 2.0 has completed its training and will be released soon. In addition, Tesla plans to begin low production of humanoid robots for internal use next year.
The rapid progress by Musk and xAI demonstrates their ambition to lead in the AI industry. If successful, xAI could soon solidify its position as a top competitor in the space, potentially overshadowing other AI startups with their advanced and powerful AI systems.
🌱 Personal Growth
How to Achieve Your Goals by December
Are you on track with your goals for 2024? If not, you’re not alone. Studies show that 95% of people who set yearly goals abandon them after just one month. We’re already more than halfway through 2024, but it’s not too late to turn things around. Whether you’re aiming to improve your career, health, or personal growth, my new video offers actionable tips and motivation to help you finish the year strong.
Don’t miss out—watch now and get back on track to achieve your goals! Link below.
🎁 Tools of the Week
Cleeai: AI search engine for professionals.
Flowtunes: DIY AI music for productivity.
Pickaxeproject: Create and embed GPT-4 apps on your website in minutes.
These were my top insights from this week. Stay tuned for more content on tech, AI, no-code, and entrepreneurship. If you liked this edition, do let me know on X/Twitter. I also share video content on my Instagram page.