• Bits & Neurons
  • Posts
  • NVIDIA is developing Arm-based processors for Windows PCs

NVIDIA is developing Arm-based processors for Windows PCs

No fluffing, busy day - in todayโ€™s edition (24 Oct):

๐Ÿ“ฐ Quick News - NVIDIA is developing Arm-based processors for Windows PCs, and more

๐Ÿ“š Nerd section - BOSS - Bootstrap Your Own Skills, and more

๐ŸŒ Web section - 5 slightly less attention-grabbing news

๐Ÿ“ฐ Quick News ๐Ÿ“ฐ

๐Ÿ‘จโ€๐ŸŽจ New technology called Nightshade enables artists to 'poison' images used for AI training without noticeable changes. These manipulated images, when used in sufficient volume, can cause an AI model's outputs to become nonsensical, disrupting the proliferation of AI art generators.

โš™๏ธ Snapdragon 8 Gen 3 chip (leaked) is set to make artificial intelligence a lot more exciting on Android phones. It promises enhanced camera capabilities, a significant speed boost, and more gaming options. Plus, your 5G experience could really start feeling next-gen.

๐Ÿ–ฅ๏ธ Nvidia, known for leading in AI computing chips, is now giving Intel a run for their money in the world of personal computers. By creating Arm-based CPUs for Windows by 2025, we might see a faster and more efficient computer performance.

๐Ÿ“œ China lodged 29,853 AI patents in 2022, 80% more than the US. This lead, established since 2017, places China as a dominant player in critical technology areas, including semiconductor production, space science, and military tech.

๐Ÿ’ฐ Microsoft is set to expand its technological presence in Australia, with plans to invest $3.2 billion over two years. This will cover artificial intelligence and cloud computing enhancements, skills training, increased cyber security and preparation for impending AI regulations.

๐Ÿง‘โ€๐Ÿคโ€๐Ÿง‘ AI โ€˜Godfathersโ€™, Bengio and Hinton, team up with 22 AI experts, suggesting a responsible framework for AI development. They suggest investing a third of AI budgets into managing possible risks, marking a pioneering step towards a safer digital future.

๐Ÿ“š Nerd section ๐Ÿ“š

๐Ÿ–ฅ๏ธ Meet BOSS โ€“ it's a major advancement in autonomous skill-learning. Using large language models, BOSS can create a multifaceted set of abilities for handling complex tasks autonomously, outperforming regular methods in unfamiliar situations.

โš’๏ธ LangChain's latest introduction, LangServe, streamlines the transition from prototypes to production versions. The tool offers various features including streaming and async support, parallel execution optimization, and input/output schemas.

๐ŸŽž๏ธ SimVP: Simpler yet Better Video Prediction. It utilizes a pure convolutional operator setting, diverging from the common attention mechanism practice. It presents a CNN-CNN-CNN architecture that not only simplifies the model but delivers significantly with surprising efficiency and reduced complexity.

๐ŸŒ Web section ๐ŸŒ 

๐Ÿ•ธ๏ธ What's this new section about?

On days when there's a lot of news, we aim to keep our newsletter short and sweet.

So here (but not in the email version) we're adding AI-related stories that may be slightly less attention-grabbing, but still could move the needle in specific sectors and fields.

๐ŸŽถ The latest update from YouTube introduces an AI-based feature for customizing playlist artwork. You could select a theme, and the system will generates appropriate imagery, providing a unique and personalized experience.

๐Ÿ’ธ Databricks is purchasing Arcion Labs, whose tech ingests data from multiple databases and SaaS applications into the Databricks Lakehouse Platform. The move ensures more informed decision-making and adherence to compliance rules.

๐Ÿ’ฒ The startup MangoBoost, which develops data processing units, raised $55 million in Series A funding. Their technology, aimed at offloading workloads from pricier CPUs and GPUs, could cut costs and improve productivity in data centers, particularly beneficial in the face of the global chip shortage.

๐Ÿง’ PimEyes, a facial recognition search engine, now safeguards childrenโ€™s digital footprint by blocking their faces from search results. This move is an effort to improve online privacy, although the implemented AI system still requires further development to accurately detect ages and angles.

Thank you for reading todayโ€™s edition!

We love to hear back from you!

Feel free to reply to our emails with questions, suggestions, or topics you'd like to see covered, or drop us a message on Twitter or Facebook.

Until tomorrow,
- Ts (Bits and Neurons)