Nvidia's New Model, Google’s Big Moves & More

In partnership with

Hello AI Enthusiasts,

Welcome to another exciting edition of The AI Pro Max! We have some groundbreaking stories that you won't want to miss.

Max Your AI Insights Today:

  • 💥 Nvidia just dropped a new AI model that crushes OpenAI’s GPT-4

  • 🛍️ Google Shopping’s getting a big AI transformation

  • 🎙️ Illuminate is a new AI podcasting tool from Google

  • 🔍 OpenAI Rolls Out SearchGPT To More Users

  • 💾 Samsung Develops Industry’s First 24Gb GDDR7 DRAM for Next-Generation AI Computing

  • 💻 Byte-Sized Buzz

  • 📚 Must-Reads

  • 🛍️ Nerdy Necessities

  • 🧠 Tech Trivia

A Shoutout to Our Sponsor

Receive Honest News Today

Join over 4 million Americans who start their day with 1440 – your daily digest for unbiased, fact-centric news. From politics to sports, we cover it all by analyzing over 100 sources. Our concise, 5-minute read lands in your inbox each morning at no cost. Experience news without the noise; let 1440 help you make up your own mind. Sign up now and invite your friends and family to be part of the informed.

  • Google Illuminate is an AI-driven podcast generator focused solely on creating podcasts, particularly from scientific research. It utilizes Google's Gemini AI model, which has a large context window of 2 million tokens, allowing it to handle extensive data more effectively than other AI models.

  • Currently, Google Illuminate is available only to a select group with a waitlist, and it is restricted to processing scientific research from arxiv.org, a free and open-access science portal run by Cornell University. This focus differentiates it from NotebookLM, another AI tool from Google.

  • Users can feed the AI with topics, PDF documents, or web addresses and adjust output settings to produce podcasts tailored for different audiences and lengths. The technology has the potential to evolve into a global AI-powered knowledge resource, capable of processing various data formats and outputting results in multiple formats, such as video documentaries or audio podcasts.

  • SearchGPT, developed by OpenAI, introduces a conversational search experience, allowing users to interact with the search engine in a dialogue-like manner. It provides direct answers to questions by summarizing key points from relevant websites, rather than just listing links. This approach includes in-line citations and a sidebar with links to the original sources for transparency and to combat misinformation.

  • SearchGPT is the result of a collaboration between OpenAI and Microsoft, leveraging Bing’s search data and OpenAI’s AI technology. This partnership enables SearchGPT to access up-to-date web information and tap into Bing’s vast index of websites, ensuring accurate and current answers to user queries.

  • SearchGPT challenges Google’s dominance in the search market by offering a more intuitive and efficient search experience. The tool impacts publishers and content creators through licensing agreements, where publishers are paid when their content appears in search results. However, it also raises technical and ethical considerations, such as the potential for inaccuracies, content valuation shifts, and the need for constant updates to maintain relevance and accuracy.

  • Google Shopping has been rebuilt from the ground up with AI, utilizing Google's Gemini models to pair with 45 billion product listings. This integration provides a more personalized shopping experience, where users receive AI-generated briefs highlighting the top things to consider for their search, along with recommended products that meet their specific needs.

  • The updated Google Shopping features dynamic filters that allow users to narrow down search results based on specific attributes such as size, availability, and location. Additionally, the platform includes categories to organize products, making it easier for users to find what they are looking for. The AI also incorporates virtual try-on and AR shopping tools to enhance the shopping experience.

  • The new Google Shopping home page includes a personalized feed populated with shoppable products and videos based on user preferences and past searches. Users can also browse a dedicated and personalized "Deals" page for curated discounts. The platform allows users to pick up where they left off in their last shopping session and provides options to turn off personalized results in the settings.

  • Samsung has developed the industry's first 24-gigabit (Gb) GDDR7 DRAM, offering the highest capacity and speed in the market. The GDDR7 achieves speeds of up to 42.5 Gbps, a 25% improvement over the previous version, thanks to the use of three-level Pulse-Amplitude Modulation (PAM3) signaling.

  • The 24Gb GDDR7 is built on a fifth-generation 10-nanometer (nm)-class process, increasing cell density by 50% while maintaining the same package size as its predecessor. It also features enhanced power efficiency, with over 30% reduction in power consumption, achieved through technologies like clock control management and dual VDD design.

  • This new DRAM is designed for high-performance applications such as data centers, AI workstations, graphics cards, gaming consoles, and autonomous driving systems. Validation with major GPU customers will begin this year, and commercial production is slated for early next year.

  • Nvidia has released the NVLM 1.0 family of multimodal large language models, with the flagship model NVLM-D-72B, which boasts 72 billion parameters. This move is significant because Nvidia is making the model weights and training code publicly available, breaking the industry trend of keeping advanced AI models proprietary. This open-source approach grants developers and researchers unprecedented access to powerful AI technology.

  • The NVLM-D-72B model demonstrates exceptional performance in both vision-language and text-only tasks, rivaling top proprietary models like OpenAI's GPT-4 and Google's advanced AI offerings. It shows improved adaptability in interpreting complex inputs such as images and memes, and it achieves a 4.3-point accuracy increase on key text benchmarks after multimodal training. The model can interpret data in charts and tables, understand memes, analyze images, and solve complex math equations.

  • Nvidia's decision to open-source NVLM 1.0 could accelerate AI research and development, enabling smaller organizations and independent researchers to contribute more significantly to the field. However, it also raises important ethical considerations, such as the potential for misuse and the need for responsible AI practices. The release may prompt industry-wide reflection on balancing innovation with accountability and could reshape the AI industry dynamics by challenging the dominance of proprietary models.

📖 Mastering the Data Paradox: Key to Winning in the AI Age

"Mastering the Data Paradox: The Key to Winning" is a compelling read that delves into the complexities of data management and utilization in today's fast-paced world. The author effectively unpacks the paradox of having vast amounts of data yet struggling to derive meaningful insights from it. The book is well-structured, offering practical strategies and real-world examples that make the concepts accessible. I particularly appreciated the actionable tips that can be implemented immediately, making it a valuable resource for both beginners and seasoned professionals.

🛒 Thames & Kosmos Kai AI Robot

Dive into the fascinating world of AI and robotics by putting together KAI, the artificial intelligence robot. Engineering geeks will be able to build their very own six-legged robot that’s able to walk around, dance, and learn from your gestures.

In what year was the first version of GPT released?

Login or Subscribe to participate in polls.

Made it to the end? Awesome! Let’s keep in touch on Twitter.

See you in our next edition!😎

Farhan

We'd love to hear your thoughts on today's email!

Your feedback helps us improve our content

Login or Subscribe to participate in polls.