• Weekly AI News
  • Posts
  • NotebookLM new features 📝, Llama-3.1-Nemotron-70B-Instruct 🦙, Meta AI CoTracker3 📹,

NotebookLM new features 📝, Llama-3.1-Nemotron-70B-Instruct 🦙, Meta AI CoTracker3 📹,

Google's latest update to NotebookLM introduces new features, including Audio Overview controls, CoTracker introduces a new point tracking model and semi-supervised training method that uses real videos, & More...

In partnership with

This year's hottest gift for businesses is BILL

Beat the Black Friday rush and get your business what it really wants for the holidays—streamlined financial operations with BILL Spend & Expense.

Choose BILL to help your business:

  • Control spending with customizable budget controls.

  • Get real-time visibility into your finances.

  • Automate expense reports to save time and focus on strategy.

Take a demo and get what YOU really want for the holidays—choose from a Nintendo Switch, Apple AirPods Pro, Samsung 50" TV, or Xbox Series S1 .

1 Terms and Conditions apply. See offer page for more details.
BILL Divvy Card is issued by Cross River Bank, Member FDIC, and is not a deposit product.

💬 OpenAI Developments

OpenAI launches dedicated ChatGPT app for Windows

OpenAI has launched a dedicated ChatGPT app for Windows, offering users a smoother experience and key functionalities like file and photo uploads, document summarization, and DALL-E 3 image generation. While still in development, the app lacks some features such as voice support and GPT Store integrations, but promises further updates later this year to enhance AI-driven productivity on desktop devices.

Bain & Company announces expanded partnership with OpenAI to accelerate delivery of AI solutions and meet fast-growing client needs

Bain & Company has expanded its partnership with OpenAI to accelerate the delivery of AI solutions, launching a dedicated OpenAI Center of Excellence. This collaboration aims to enhance AI capabilities for Bain’s clients, particularly in industries like retail and healthcare, by leveraging OpenAI's advanced technologies, including multi-modal and real-time applications, to drive business innovation and efficiency.

🌐 AI in Various Sectors

Healthcare: Google's new Therapeutic Large Language Model (Tx-LLM) is designed to accelerate drug discovery by predicting key drug properties, signaling a shift toward industry-specific AI solutions. This model, tailored for pharmaceutical research, highlights the growing trend of using specialized AI to address complex challenges in sectors like healthcare, manufacturing, and finance, offering improved efficiency and adaptability across industries.

Robotics: Saudi Arabia is making significant advances in AI-driven healthcare robotics, highlighted by the world's first fully robotic heart transplant performed at King Faisal Specialist Hospital. This achievement demonstrates the potential of AI and robotic surgery to enhance precision, reduce recovery times, and minimize risks, positioning Saudi Arabia as a leader in healthcare innovation.

Social Media and Interaction: Adobe's Project Super Sonic, revealed at the 2024 MAX conference, uses AI to generate sound effects for videos through text prompts, object recognition, and user-generated vocal sounds. This tool offers creative control and may eventually be integrated into Adobe's software, though it's currently a prototype.

🚀 Tech Industry Moves

Anthropic has released a major update for its Claude AI iOS and Android apps, introducing new customization features, the ability to search past chats, and an iPad app debut, enhancing user productivity. These updates aim to improve accessibility and user experience, positioning Claude AI as a versatile tool for personal and professional use across devices.

Mistral AI has launched two new language models, Ministral 3B and Ministral 8B, designed for edge devices like smartphones and laptops, offering powerful AI capabilities without relying on cloud infrastructure. This shift towards edge computing could enhance privacy, reduce latency, and make advanced AI more accessible while addressing environmental concerns.

Meta's chief AI scientist Yann LeCun believes achieving human-level AI will require building "world models" that understand and interact with the three-dimensional world, a concept that could take a decade to realize. While today's AI models are good at predicting text and pixels, they lack the ability to reason, plan, and intuit in the way humans do, making world models the next frontier for advancing AI capabilities.

Meta researchers, in collaboration with UC Berkeley and NYU, have developed a new method called "Thought Preference Optimization" (TPO) to enhance AI models' reasoning capabilities. TPO improves performance by having AI generate internal thoughts before answering, allowing more effective problem-solving across various tasks, although performance declined in highly specialized areas like math.

⚖️ AI Ethics and Regulation

Why corporate integrity is key to shaping the use of AI? Corporate integrity is crucial for ensuring responsible AI development, as global regulations struggle to keep pace with rapidly advancing technology. Businesses can lead by integrating AI ethics into their governance structures, supported by ethical guidelines, due diligence, and comprehensive risk management systems. Boards and investors also play a key role by fostering AI accountability, with rising shareholder activism pushing for ethical AI practices. Emphasizing corporate integrity helps build trust, mitigate risks, and position ethical AI as a strategic advantage for long-term success and societal benefit.

💡 Emerging AI Technologies and Companies

CoTracker3: Simpler and Better Point Tracking by Pseudo-Labelling Real Videos

CoTracker introduces a new point tracking model and semi-supervised training method that uses real videos without annotations, generating pseudo-labels through off-the-shelf models. This approach simplifies existing tracking models, requiring significantly less data while outperforming recent trackers on standard benchmarks. CoTracker reliably tracks visible and occluded points, achieving impressive results even when points are occluded or leave the field of view.

Nvidia's new open-source AI model beats GPT-4o

Nvidia has quietly launched a new AI model, Llama-3.1-Nemotron-70B-Instruct, which outperforms OpenAI's GPT-4 and Anthropic's Claude 3.5 in benchmark tests, marking a significant shift in its AI strategy. By refining Meta's Llama 3.1 model using advanced techniques like Reinforcement Learning from Human Feedback, Nvidia offers a high-performing, customizable AI solution. This release highlights Nvidia's transition from a GPU leader to a major AI software contender, potentially reshaping the competitive landscape and pushing rivals to accelerate innovation.

🎨 AI in Creative Industries

Google's latest update to NotebookLM introduces new features, including Audio Overview controls, allowing users to generate AI-hosted audio summaries based on uploaded sources. Additionally, Google announces NotebookLM Business, an upcoming version for organizations, offering enhanced features and data security through Google Workspace. Users can apply for early access to the pilot program, with more details on general availability and pricing expected later this year. NotebookLM already serves over 80,000 organizations.

✨✨✨✨

Andreessen Horowitz discusses how AI-powered "wedges"—innovative, narrowly-focused products—can help startups disrupt industries traditionally dominated by established players. By leveraging large language models (LLMs), startups can capture and structure data in ways that were previously impossible, outpacing incumbents in areas like sales, insurance, and IT service management. These AI-driven tools, when integrated with legacy systems, enable faster and more efficient data collection, providing startups with an edge in transforming "slow boil" industries into rapidly growing markets.

🌍 Global AI Race

Head of Saudi tech institute pledges to limit China AI collaboration

The new head of Saudi Arabia's King Abdullah University of Science and Technology (Kaust), Professor Sir Edward Byrne, has pledged to limit AI collaboration with China to avoid jeopardizing access to US-made AI chips. Byrne emphasized prioritizing partnerships with the US, UK, and Europe, as the US tightens export controls on AI technology amid growing concerns in the Middle East about chip access.

If you want more updates related to AI, subscribe to our Newsletter

👥 Connect & Feedback!

👉 Join Us:

📧 Advertise In Weekly AI News:

📧 Contact directly at [email protected]

😍 Share with your friends!


Reply

or to participate.