OpenAI Enhances ChatGPT with Memory Feature, Raises Privacy Concerns
OpenAI has introduced a new memory feature for its ChatGPT, allowing the AI-powered bot to retain information about users and their conversations. This enhancement can be utilized in two modes: users can instruct ChatGPT to remember specific details or allow the bot to autonomously gather information during conversations. The memory feature is now available across all variants of ChatGPT, significantly enhancing user experience in various domains such as education, travel services, and fitness apps. However, this development has raised concerns about user privacy.
To address these concerns, OpenAI emphasizes user control over ChatGPT’s memory. Users have the ability to review the stored information, request data removal, or activate the Temporary Chat mode for anonymous conversations. OpenAI has also introduced a Manage Memory section within the settings, providing granular control over stored data to prioritize user privacy.
EPIC Foundation Introduces India’s First AI-Powered Milkyway Tablet
EPIC Foundation, in collaboration with Technologies, MediaTek India, and CoRover.ai, has unveiled the Milkyway Tablet, India’s first AI-driven education tablet. This tablet integrates the BharatGPT virtual assistant, offering digital learning and language translation capabilities. The key focus of this initiative is to bridge the digital divide and address e-waste concerns. The Milkyway Tablet features an 8-inch 1280×800 IPS display, MediaTek MT8766A processor, 3/4GB RAM, 32/64GB storage, 5MP front and 8MP rear cameras, and a 5,100mAh battery. Powered by Android 13, it supports 4G LTE connectivity and includes features like video and chatbots, as well as a speech translation app. The goal of this initiative is to provide sustainable and affordable education solutions while fostering demand for locally-made electronics.
Nvidia Launches ‘Chat with RTX’ Chatbot on PC, Leveraging Locally Stored Content
Nvidia has launched an early version of ‘Chat with RTX,’ an innovative personal AI chatbot that allows users to customize it using their own content on Windows PCs. This free tech demo harnesses local files as sources of knowledge, enabling users to create a chatbot by adding datasets from documents and notes stored on their computers. ‘Chat with RTX’ utilizes retrieval-augmented generation (RAG), Nvidia TensorRT-LLM software, and Nvidia RTX acceleration, providing generative AI capabilities to local, GeForce-powered Windows PCs. Distinguishing itself, this chatbot supports various file formats and seamless integration with multimedia sources like YouTube videos. It operates on Windows RTX PCs and workstations, offering fast responses while ensuring data privacy by handling sensitive information locally without the need for internet connectivity.
Singapore to Develop Localized AI Models, Introducing SEA-LION
Singapore is striving to scale its Southeast Asian-focused artificial intelligence (AI) models for commercial use by 2024, positioning itself as a hub for AI technology. Leading this effort is AI Singapore, a government-led initiative, which plans to develop its own AI models tailored to the local populations. One notable model under development is SEA-LION, designed as an open-source large language model (LLM) similar to ChatGPT, specifically representing the languages and cultures of Southeast Asia. SEA-LION is expected to be trained on data in 11 languages, including Vietnamese, Thai, and Bahasa Indonesia, and is projected to be completed by 2024. Leslie Teo, senior director of AI products at AI Singapore, has revealed plans to expand SEA-LION’s capabilities to encompass multimodal speech-text models, with a focus on growing its corpus to 2-3 trillion and expanding its parameters to 30-50B this year.
SEC Chairman Warns Companies Against Exaggerated AI Claims
Gary Gensler, the head of the US Securities and Exchange Commission (SEC), has cautioned publicly traded companies against AI washing and emphasized the importance of transparency when discussing artificial intelligence with investors. Gensler urges corporations to provide clear explanations of their AI usage, including operational risks, and consider disclosing executives’ remarks on the technology to investors. With an increasing number of SEC registrants highlighting AI in their annual reports, Gensler stresses the significance of adhering to securities laws. While acknowledging the transformative potential of AI, he also highlights concerns about its implications for financial stability. Gensler warns of systemic risks arising from the widespread use of AI models and underscores the need for regulatory oversight to address biases and concentration among AI providers.
The introduction of OpenAI’s memory feature in ChatGPT enhances user experience. However, it raises privacy concerns, which OpenAI addresses by offering users control over the bot’s memory. EPIC Foundation’s Milkyway Tablet aims to provide sustainable and affordable education solutions in India. Nvidia’s ‘Chat with RTX’ allows users to customize a personal chatbot using locally stored content. Singapore plans to develop localized AI models, focusing on Southeast Asian cultures and languages. SEC Chairman Gary Gensler advises companies to be transparent and cautious when discussing AI with investors. The increasing adoption of AI necessitates adherence to securities laws and regulatory oversight to mitigate potential risks.
