Global AI Weekly
Issue number: 37 | Tuesday, January 30, 2024
Highlights
}
Satya Nadella says the explicit Taylor Swift AI fakes are ‘alarming and terrible’
Microsoft could improve its image generator guardrails.
theverge.com
}
New embedding models and API updates
We are launching a new generation of embedding models, new GPT-4 Turbo and moderation models, new API usage management tools, and soon, lower pricing on GPT-3.5 Turbo.
openai.com
}
Reasoning and reliability in AI
PhD students interning with the MIT-IBM Watson AI Lab look to improve natural language usage.
news.mit.eduVideo
}
Transform RAG and Search with Azure AI Document Intelligence
Retrieval Augmented Generation (RAG) is a design pattern commonly used in Document Generative AI (e.g., chat with your document). Semantic chunking is a key step in RAG to get efficient storage and retrieval. The session will introduce the latest capabilities provided by Azure AI Document Intelligence to enable semantic chunking and demonstrate a RAG application with semantic chunking for improved accuracy. We will also show how to get started using Azure AI Document Intelligence to build your own RAG application.
youtube.com
}
Process Automation with LLMs and Prompt Flow
Explore Process Automation with LLMs and Prompt Flow
youtube.comArticles
}
Taylor Swift deepfake pornography sparks renewed calls for US legislation
Fake but convincing explicit images of pop singer were viewed tens of millions of times on X and Telegram, prompting outcry from US politicians The rapid online spread of deepfake pornographic images of Taylor Swift has renewed calls, including from US politicians, to criminalise the practice, in which artificial intelligence is used to synthesise fake but convincing explicit imagery.
theguardian.com
}
MetaOpt: Examining, explaining, and improving heuristic performance
MetaOpt helps analyze, explain, and improve heuristic performance before deployment in production systems. Learn how it works, particularly in traffic engineering, packet scheduling, and VM placement.
microsoft.comGetting Started With CUDA for Python Programmers
if, like me, you've avoided CUDA programming (writing efficient code that runs on NVIGIA GPUs) in the past, Jeremy Howard has a new 1hr17m video tutorial that demystifies the basics. The code is all run using PyTorch in notebooks running on Google Colab, and it starts with a very clear demonstration of how to convert a RGB image to black and white.
simonwillison.net
}
Accelerating Generative AI with PyTorch IV: Seamless M4T, fast
This post is the fourth part of a multi-series blog focused on how to accelerate generative AI models with pure, native PyTorch. To skip to the code, check out our github (seamless_communication, fairseq2). We are excited to share a breadth of newly released PyTorch performance features alongside practical examples to see how far we can push PyTorch native performance. In part one, we showed how to accelerate Segment Anything over 8x using only pure, native PyTorch.
pytorch.org
}
Bard Advanced
Google introduces "Bard Advanced," a groundbreaking artificial intelligence chatbot driven by Gemini Ultra, the pinnacle of their large language framework. Currently, developers enjoy complimentary access to Gemini Pro and Gemini Pro Vision through Google AI Studio, accommodating up to 60 queries per minute, catering to diverse app development requirements.
medium.comGlobal Context Vision Transformers
We propose global context vision transformer (GC ViT), a novel architecture that enhances parameter and compute utilization for computer vision. Our method leverages global context self-attention modules, joint with standard local self-attention, to effectively and efficiently model both long and short-range spatial interactions, without the need for expensive operations such as computing attention masks or shifting local windows.
research.nvidia.com
}
What to do about AI in health?
Although artificial intelligence in health has shown great promise, pressure is mounting for regulators around the world to act, as AI tools demonstrate potentially harmful outcomes.
news.mit.edu
}
The Rabbit R1 will receive live info from Perplexity’s AI ‘answer engine’
Perplexity's pplx-online LLM APIs, will power Rabbit R1 for providing live up to date answers without any knowledge cutoff! CES 2024 darling Rabbit has announced a partnership with Perplexity that will link the “conversational AI-powered answer engine” to the R1, a $199 Teenage Engineering-designed AI gadget that’s already sold through 50,000 preorders.
theverge.comUpcoming Events
}
Global AI Bootcamp 2024 | March 2024
The Global AI Bootcamp is an annual event that occurs worldwide, where developers and AI enthusiasts can learn about AI through workshops, sessions, and discussions. Local chapters of the Global AI Community host the events in various locations across the globe. This year, the bootcamp will take place in the full month of March 2024.
globalai.community