Spotlight

1. Tom Hanks raises concern over unauthorized AI version of himself being used in an advertisement.
2. Meta introduces Llama 2 Long, an improved AI model that outperforms competitors.
3. Spotify launches an AI voice translation tool for podcasts.
4. Getty Images unveils an AI-powered image generator to prevent copyright infringement.
5. Researchers develop DistTGL, a method for training Memory-based Temporal Graph Neural Networks (M-TGNNs).

MindBlowing AI Tools Everyone talking about

– Meta’s Llama 2 model is now integrated into various platforms (Instagram, WhatsApp, and Facebook Messenger) as Meta AI, offering real-time information and image generation capabilities.
– OpenAI upgrades ChatGPT to support voice commands and picture uploads, expanding beyond text-only prompts.
– SAP introduces Joule, a generative AI copilot for its enterprise portfolio, enhancing user experience and decision-making processes.
– Apple plans to bring its powerful internal search engine, Pegasus, to the App Store and other apps, potentially utilizing generative AI tools.

What The World Have Researched & Innovated

Researchers from Soochow University contribute OpenBA-15B, a 15B bilingual seq2seq model, aiding Chinese-centric NLP advancements.
Vision transformers face challenges with recycling useless patches, but the addition of dedicated register tokens resolves the issue.
Microsoft researchers introduce AutoGen, an AI framework for simplifying LLM workflows, along with StoriaBoard, a pre-visualization tool.
VidChapters-7M, a dataset of segmented videos, is released for training and benchmarking in video chapter generation tasks.
AI chatbots raise questions about the reliability of testimony for acquiring knowledge.

Stay updated, Let’s Know Together.