Significant progress in mobile network evolution, focusing on the shift to 6G. Generative AI optimizes networks, enables dynamic spectrum sharing, intelligent network slicing, predictive maintenance, enhanced security, edge computing, and AI-driven communication in 6G. Synergy between generative AI, 5G, and 6G shaping future telecom.

Salesforce enhances AI products with vector database support, AI search for Copilot. Data Cloud integrates unstructured data into CRM workflows for powerful insights. Salesforce aims to launch these features in February 2024.

SenseTime's founder Tang Xiaoou, an AI pioneer, has passed away at 55. Uncertainty looms over the company's future amid challenges in regulation, investment, and market position. Tang's legacy and vision will be crucial as SenseTime navigates industry changes.

MIT researchers address image difficulty in object recognition, introducing minimum viewing time (MVT) metric. Their study reveals challenges in traditional datasets and models. MVT enhances object recognition evaluation and aims for human-like performance.

Earlier this year, Dropbox introduced new AI-driven search tools for paid customers. However, to test Dropbox's AI, users must allow OpenAI access to their files. This has raised concerns among users about data privacy. The feature can be enabled in settings, but users have reported it being activated without their knowledge. Users can manage AI features, but caution is advised due to data sharing concerns. Dropbox CEO Drew Houston clarified that data is not passively sent to OpenAI.

Meta is enhancing Instagram by introducing new AI features like generative text suggestions for background editing, accessible to US users. The parent company aims to improve user experience on Facebook, Instagram, Messenger, and WhatsApp with over 20 new generative AI-driven tools in 2023.

Miniature robots play crucial role in exploring tight spaces, like Mugatu, the manipulable bipedal robot developed by Aaron Johnson and Sarah Bergbreiter's team. Simplified walking mechanism shows promising potential for future autonomous robots.

Instagram introduces new generative AI-powered background editing tool, allowing users to change backgrounds in stories with clever prompts. Meta also unveils AI plans, including 28 AI-powered characters on WhatsApp, Messenger, and Instagram in the US.

Stability AI launches membership system with three levels: non-commercial (free), professional ($20/month), and enterprise (custom pricing). Aim to make AI tech accessible and affordable, offering flexibility and control over data. CEO Emad Mostaque resigns amid financial challenges.

OpenAI plans to develop AI systems surpassing human capabilities, prompting concerns of potential risks. They partner with Eric Schmidt for $10 million "Superalignment Fast Grants" to ensure future AI safety and alignment. Research focuses on controlling superhuman systems for human benefit, leveraging top talents across various fields to address crucial technological challenges.

Google launches AI models for healthcare sector named MedLM, based on Med-PaLM. Models can answer medical questions, write summaries, passed US Medical Licensing Exam, and customized. Aim to enhance patient outcomes and operator efficiency.

Alibaba Group researchers develop Qwen-Audio, a versatile audio-language model excelling in speech perception without fine-tuning. Qwen-Audio-Chat supports multi-turn conversations, showcasing powerful audio understanding. Future Qwen-Audio enhancements aim to improve performance on diverse tasks and languages, addressing interference through refined training methods.

OpenAI partners with Axel Springer to integrate news from Politico & Business Insider into ChatGPT for model training. Collaboration may help OpenAI address EU concerns on privacy & data usage. Grok's real-time capabilities also challenge ChatGPT.

Microsoft open-sources AI model Llama 2, a Meta Platforms competitor, as Model as a Service in Azure AI Studio. Customers can now access various Llama models for AI tasks through Azure's cloud platform, providing an alternative to OpenAI's GPT models. OpenAI's GPT-4 Turbo with Vision is also available on Azure AI Studio for analyzing visual content.

ByteDance used OpenAI's tech to secretly develop a language model, violating terms of service. Account suspended by OpenAI. Employees tried to hide evidence and violated Microsoft regulations. OpenAI confirmed suspension, while Microsoft emphasized responsible tech usage.

The CEO of Ola, Bhavish Aggarwal, launched Krutrim, India's first full-stack AI. Trained on 20 trillion tokens, it supports 20 Indian languages, outperforming GPT-4. Krutrim Pro, a multimodal AI, is set to release next quarter. However, key details like training dataset and team remain undisclosed.

Bitdefender introduces Scamio, a free online tool using AI to detect scams through text, images, and links. Users can interact with it through a web browser or Facebook Messenger, ensuring a safer way to verify suspicious content. Trust in this AI detector is supported by Bitdefender's long-standing reputation in cybersecurity.

Google released Duet AI for Developers, an AI-powered code completion tool. Partnered with 25 companies to provide datasets for platform-specific code writing and problem-solving. Free until Jan 2024, then $19/user/month.

Some users report GPT-4's "laziness" as winter nears, sparking speculation of seasonal behavioral changes impacting AI model behavior. OpenAI recognizes the issue but hasn't updated the model. Discussions around the "winter break hypothesis" continue, emphasizing the unpredictable nature of large language models like GPT-4.

Google launches MusicFX, an AI-driven music generation tool using MusicLM and SynthID. Limitations include no artist-specific music. Engage with AI responsibly through Google's Test Kitchen, prioritizing public feedback. Privacy protected, data stored anonymously. MusicFX impacts music creation and industry, raising ethical and copyright concerns.