Mohammad Nouman Khan’s Post

🚀 Breaking: Meta just unleashed Llama 3.2, and it's not just an upgrade - it's a revolution in AI accessibility! Ever dreamed of running state-of-the-art AI on your smartphone? Or deploying vision models that rival the big tech giants? Llama 3.2 makes it a reality: => Tiny but mighty: 1B and 3B models fit on edge devices, supporting 128K token context => Vision powerhouses: 11B and 90B models outperform closed competitors like Claude 3 Haiku => Ecosystem ready: Optimized for Qualcomm, MediaTek, and Arm processors =>Llama Stack: Simplified deployment for RAG and safety-integrated applications =>True openness: Available for download, fine-tuning, and immediate use on major cloud platforms This isn't just another AI release - it's Meta democratizing cutting-edge tech for developers worldwide. I decided to test the Llama 3.2 11B Vision Instruct model (for free) using OpenRouter. See a couple of tests below. What groundbreaking applications will you build with Llama 3.2? #AIInnovation #LlamaAI #OpenSourceAI #MetaAI ______________________________________ 🛠 Do you need help identifying how generative AI can enhance your product or service? I can help! https://intellegen.ai/ 📘 Get my book: Unlocking the Secrets of Prompt Engineering https://lnkd.in/ekGzCejd 🚀 Get SheetSmart, a FREE Google Sheets template with the power of ChatGPT https://lnkd.in/emXPW59p

The transformative nature of AI I've been immersed lately in the realm of AI and workflow and it has made me contemplate the historic point of time in which we stand. Historical moments in time that changed the world: The Gutenberg Press 1450s The Industrial Revolution The invention of Personal Computers The Internet The iPhone and Smartphones The Cloud AI in the form of ChatGPT or Microsoft's Copilot Every time I use CoPilot to streamline research, to summarize notes or to create Presentations I am reminded of just how transformative the technology is. It truly is an amazing time to be alive. For some the sun is setting as we leave the past behind. For others it is a new day and the sun is rising with unimagined possibilities about to be unlocked. Either way it's a moment to stop, contemplate the epic change and step confidently into a new day with new possibilities. https://msft.it/6046ciLKM #microsoft #copilotformicrosoft365

Interesting AI news from across the web: 🤖 Wallow is taking Slack to the next level by tailoring it specifically for makers. [Read More](https://lnkd.in/gmd9kzsu) 📡 NVIDIA’s big move: Deploying AI tech in Ooredoo datacenters across five Middle Eastern countries. [Read More](https://lnkd.in/gW2sNiTh) 💬 Meta just launched its AI chatbot in India, but it currently only supports English. [Read More](https://lnkd.in/g-aN2P7f) 🚀 Qualcomm makes its AI models available for developers to build new AI-enabled apps. [Read More](https://lnkd.in/g45PGVJw) 🛠️ Anika Collier Navaroli is working to shift the power imbalance in AI. [Read More](https://lnkd.in/gBF9-keG)

📢 Exciting News in AI! 🚀 Meta's latest breakthrough, Llama 3.1, is here and it's setting new benchmarks in the world of artificial intelligence. With a staggering 405 billion parameters and trained using over 16,000 Nvidia H100 GPUs, Llama 3.1 outperforms even the top proprietary models like OpenAI's GPT-4 and Anthropic's Claude 3.5 Sonnet in certain areas. Meta's ambitious goal is to make Llama 3.1 the most widely adopted AI assistant by the end of the year, even surpassing the popular ChatGPT. One of the most exciting aspects is the model's ability to generate images based on a user's likeness, expanding its usability across more countries and languages. Mark Zuckerberg envisions a future where open-source AI models, like Llama 3.1, will dominate, much in the way that Linux became the go-to open-source operating system. For those eager to leverage this powerful tool, Llama 3.1 is readily available for download and can be accessed via cloud services like AWS, Azure, and Google Cloud. This marks a significant leap in the AI landscape, offering innovative solutions and broadening the accessibility of advanced AI tools. Stay tuned and get ready to experience the future of AI with Llama 3.1! #AI #Meta #OpenSource #Innovation #DigitalTransformation #MachineLearning #Llama3.1 #ArtificialIntelligence #TechNews #MarkZuckerberg

today in AI: 1/ Microsoft has announced a copilot for Telegram. Meta's got AI on WhatsApp, Instagram, and Facebook, but Microsoft isn't about to be left behind in the AI assistant race. No social media platform? No problem! Microsoft's bringing the AI magic of Copilot to Telegram. Our summary: https://lnkd.in/d84eYQfy 2/ Elon Musk will build a Gigafactory of Compute - Following xAI’s series B of $6B, Elon aims to connect 100,000 of Nvidia's most powerful H100 GPUs, creating a computing powerhouse four times larger than any existing AI cluster. This AI Gigafactory will power Grok and xAI’s other projects. Our summary: https://lnkd.in/dSp4EYKN #AI #Telegram #AIGigafactory

It’s been an amazing week for AI with lots of exciting updates 🔥 > OpenAI launched Advanced Voice Mode. > Google open-sourced Gemma 2 2B. > Midjourney released version 6.1. > Synthesia introduced “Personal Avatars.” > Stability AI launched Stable Fast 3D. > Vimeo now has AI video translation. > Canva acquired Leonardo AI. > GitHub launched ‘GitHub models.’ > Microsoft sees OpenAI as a competitor. > Black Forest Labs released Flux 1. > NVIDIA uses Apple Vision Pro for robot training. > Runway introduced Gen-3 image-to-video tech. > 01 AI unveiled Yi-Vision. > Meta introduced SAM 2. > Character AI open-sourced ‘Prompt Poet.’ > Friend launched an AI companion necklace. > Apple shared its Apple Intelligence report. > Perplexity announced revenue sharing for publishers. > Figure showed off the ‘most advanced’ humanoid robot. > Character AI's cofounders joined Google. #ai #genai #llms #openai #google #github #microsoft #nvidia #runway #meta #apple #perplexity

Hello fellow AI Enthusiasts... Meta has just launched Llama 3.1, the largest open-source AI model to date, boasting 405 billion parameters and trained with over 16,000 Nvidia H100 GPUs. This model is set to rival top proprietary models like OpenAI's ChatGPT. Much like ChatGPT, Llama 3.1 aims to be a versatile AI assistant, but Meta's open-source approach could democratize AI development, similar to how Linux revolutionized operating systems. Mark Zuckerberg envisions Llama 3.1 becoming the most widely adopted AI assistant by the end of the year, outpacing ChatGPT. Some pretty cool features and it will be distributed and integrated across Meta properties. I'll be trying it and reporting back. #AI #advertising #growthmarketing #newbusiness #soulpurposeadvisory

It’s March 2024, let’s quickly recap the highlights in #AI for this year! Starting with the recent #Microsoft breakthrough, their 1-bit #LLM — this allows #LLMs to run on low end devices and save tons of energy (up to 70%). We have #Groq and their new AI chip that seems to be running super fast. The chip is designed specifically to boost performance of the AI applications — named #LPU. You can check their demo online, just Google it. Heavy investment in #robotics — #Humanoid, a robotic startup has got $2.6 billion #investment from #Bezos, #MS, and #Nvidia. Meanwhile #Musk showed a new demo of the #Tesla Bot. #OpenAI shakes the world with their text to video model, #Sora — which’s still behind closed doors. Some digging has been done so far, revealing that #UE5 generated videos were used in the training as well as its architecture, which is based on rejected paper released earlier by #LeCun! Oh, and Musk sues OpenAI for derailing from its original purpose — that is benefiting humanity with AI. #Google started strong and then things went a bit twisty — their latest #Gemini model with a huge context window seems to be performing very well. Google also seems to be joining the #OpenSource again with their #Gemma models, highly censored models that are almost useless. Gemini image generation has been released and this is where things get interesting. Google tried their best to make the model “safe” and “diverse” to a point where it rewrote the history! Generating black and Asian people most of the time — literally anything but not white people. Google pulled the trigger and disabled the feature, apologized, and promised to fix it soon. If this case doesn’t show you why we need OpenSource AI, I don’t know what else does! Other updates to be aware of — #Meta teasing #Llama3, #Claude3 is out, #StableDiffusion3 is almost out, #SunoAI3 is in beta, #Vercel showcased #GenUI where you can render #UI components on the fly as you chat with an LLM — aaaand because all the above, some states in the #US and countries started testing #UBI (universal basic income), aka. taking it seriously.

Boost your productivity with contextually relevant answers: Explore Chat with RTX from #NVIDIA This innovative demo app by NVIDIA lets you build a personalized #AI assistant chatbot, trained on YOUR data - docs, notes, even videos! It leverages cutting-edge technology to answer your questions with contextually relevant insights, all stored and processed securely on your PC, without requiring an internet connection. #nvidia #chatwithrtx #ai #chatbot 🔗https://lnkd.in/dFBaNjPY

Here's this week's tech news. 1. Microsoft Copilot makes the advanced GPT-4 Turbo model free to all users. 2. Elon Musk announced that his AI chatbot Grok will become open-source. But it’s unclear which parts of Grok will be open to the public. 3. The House has passed a bill that could lead to a US ban on TikTok unless the app separates from its Chinese parent company, ByteDance. 4. Apple acquired DarwinAI, a Canadian company specializing in AI technology for manufacturing inspections and efficient AI systems. Apple is integrating generative AI into iOS 18 and macOS 15. Plans to run AI features on-device for privacy, leveraging DarwinAI’s expertise in creating compact AI systems. #techupdates #technews #copilot #applenews #darwinAI