- The AI-ronman π
- Posts
- North Korea experiments with AI cyber-attacks π€
North Korea experiments with AI cyber-attacks π€
The AI-ronman π
π Step into the AI Wonderland with us! As we push the boundaries of what's possible, we invite you to join us in uncovering the secrets, surprises, and the sheer magic of artificial intelligence. π©
Quick Takes β‘
OpenAI eyes $86 b valuation in upcoming employee share sale
Nvidia and iPhone maker Foxconn to build 'AI factories'
Anthropic faces landmark copyright lawsuit over AI chatbot training
IBM's NorthPole chip: a brain-inspired computer chip
North Korea experiments with AI cyber-attacks
Meta AI transforms brain waves into real-time images
OpenAI puts the brakes on Arrakis project
Infosys and Google Cloud: A renewed alliance for AI-First Enterprises
Azure AI content safety: Microsoft's new tool for safer cyberspace
Nvidia bets on generative AI for robotics and industrial applications
Deep Dive π
OpenAI eyes $86 b valuation in upcoming employee share sale π
OpenAI plans to sell employee shares at a massive valuation of $86 billion. The company, 49% owned by Microsoft, is poised to generate $1.3 billion in yearly revenue, largely from ChatGPT Plus subscriptions. Yet, due to the costly nature of AI, it aims to secure more funds for continued development.Nvidia and iPhone maker Foxconn to build 'AI factories' π
Nvidia and Foxconn are joining forces to build 'AI Factories' that will use Nvidia's chips for various tech applications, including self-driving cars and language AI.π€ This partnership emerges as the U.S. plans to limit chip exports to China. The collaboration not only helps Foxconn transition into a platform solutions provider but also solidifies Nvidiaβs trillion-dollar market valuation. π
Anthropic faces landmark copyright lawsuit over AI chatbot training βοΈ
Universal Music, ABKCO, and Concord are suing Anthropic for using copyrighted lyrics to train its chatbot, Claude. The lawsuit, a first against Anthropic, claims at least 500 songs' lyrics were used without permission. Financially backed by tech giants like Google and Amazon, Anthropic's case could set legal precedents for AI training. The suit seeks damages and an injunction against further infringement.
IBM's NorthPole chip: a brain-inspired computer chip π§
IBM's NorthPole brings a blend of speed and energy-efficiency to AI, excelling in tasks such as image recognition. The design of the NorthPole chip includes wired connections inspired by the connections between parts of the human cerebral cortex. It's 5 times more energy-efficient and faster than existing chips. π― If built with the latest tech, its performance could leap 25-fold. π New materials and memristors could further boost its capabilities.
North Korea experiments with AI cyber-attacks π€
North Korea is ramping up its cyber-attack capabilities by using AI, according to a US official. They are leveraging machine learning to quickly write malware and find vulnerable targets. Previous attacks like Sony and WannaCry have already caused billions in damages worldwide. π» Revenue from cyber crimes may fund North Korea's missile program, motivating continued development. π‘ Defending against AI-enabled cyber threats has become more critical than ever.
Meta AI transforms brain waves into real-time images πΌ
Meta AI has created a groundbreaking system that can generate images from brain data in real-time. π§ Using a technique called magnetoencephalography (MEG), this innovation could lead to non-invasive brain-computer interfaces. The AI system consists of an image encoder, brain encoder, and image decoder. The AI system's brain signals align well with modern computer vision AI systems like DINOv2. This not only furthers our understanding of how the brain works but could help those with speech disabilities. π€
OpenAI puts the brakes on Arrakis project π
OpenAI has shelved its Arrakis project, an endeavor aimed at creating a more efficient, GPT-4-level AI model.π» Originally built on a sparse neural network concept, the model failed to meet performance expectations. However, OpenAI is exploring its incorporation into other projects, such as the Gobi multimodal model, and is also working on a faster GPT-4 version. π
Infosys and Google Cloud: A renewed alliance for AI-First Enterprises π€
Infosys and Google Cloud are strengthening their partnership to turn businesses into AI-centric organizations. Infosys will set up Generative AI Labs and train 20,000 people on Google's advanced AI tech. The collaboration aims to improve Infosys platforms like Live Enterprise Application Management and Customer Intelligence, fostering long-term digital growth for businesses. π
Azure AI content safety: Microsoft's new tool for safer cyberspace π‘οΈ
Microsoft has launched Azure AI Content Safety for public use, aiming to help organizations keep their online platforms clean. π» Using advanced AI models, it detects and flags inappropriate content like violence. The platform is customizable, allowing businesses to set their own moderation policies to meet customer needs. π
Nvidia bets on generative AI for robotics and industrial applications π
NVIDIA is supercharging its Isaac ROS robotics framework and Metropolis platform on Jetson with advanced generative AI and cloud-native features. Developers can tap into specialized tools at the Jetson Generative AI Lab, while businesses can leverage potent APIs for vision-based applications.π» NVIDIA has also rolled out workflow templates and a new system service, JetPack 6, to simplify edge AI development. The expanded offerings aim to accelerate the adoption of AI in robotics and edge computing. π‘
π§ββοΈ As we close this AI chapter, remember: you're the wizard of your own data-driven destiny! π
Ciao for now!
(Author: Poonam π§ )
Karan π π
CEO, GlazeGPT.com