Ray-Ban Meta smart glasses being worn.
The AI landscape witnessed significant developments this week, from OpenAI’s substantial funding round to Nvidia’s surprising LLM release and a concerning privacy breach involving Meta smart glasses. This article covers five of the most impactful AI announcements.
OpenAI Secures Billions in Funding
OpenAI CEO Sam Altman on stage.
OpenAI’s recent funding round secured a staggering $6.6 billion in investments, with participation from existing investors like Microsoft and Khosla Ventures, alongside new investors like SoftBank and Nvidia. This latest round brings OpenAI’s valuation to an impressive $157 billion, solidifying its position as one of the world’s most valuable private companies. Should OpenAI’s proposed for-profit restructuring succeed, CEO Sam Altman could see his equity soar to over $150 billion, potentially placing him among the top 10 richest individuals globally. Following this funding news, OpenAI launched Canvas, a feature inspired by Anthropic’s Artifacts, designed for collaborative AI development.
Nvidia Unveils Open-Source LLM
Nvidia CEO Jensen Huang.
Nvidia expands its AI presence beyond hardware with the release of LVNM 1.0, an open-source large language model. LVNM excels in various vision and language tasks, and the flagship LVNM-D-72B boasts 72 billion parameters, rivaling the capabilities of GPT-4. However, Nvidia positions LVNM not as a direct competitor to existing leading LLMs, but as a versatile platform for developers to build their own chatbots and AI-powered applications.
Google’s Gemini Live Expands Language Support
Gemini Live on a Google Pixel 9.
Direct conversation with AI chatbots is becoming a key feature, and Google has expanded Gemini Live’s multilingual capabilities to nearly four dozen languages. This expansion includes French, German, Portuguese, Hindi, and Spanish. Microsoft also introduced a similar voice interaction feature for Copilot, named Copilot Voice, designed for intuitive brainstorming on the go. These developments join ChatGPT’s Advanced Voice Mode and Meta’s Natural Voice Interactions, enabling users to engage in more natural conversations with their devices.
California Governor Vetoes AI Safety Bill
California Governor Gavin Newsom.
California Governor Gavin Newsom vetoed SB 1047, the Safe and Secure Innovation for Frontier Artificial Models Act. Newsom expressed concerns that the bill’s focus on large language models overlooked the potential risks of smaller, specialized models, which could pose equal or greater dangers.
Meta Smart Glasses Modified for Doxing
Ray-Ban Meta smart glasses by a pool.
Harvard computer science students demonstrated a concerning vulnerability in Meta smart glasses by modifying them to identify and retrieve personal information of individuals within their field of view. The I-XRAY experiment uses PimEyes image recognition software to identify individuals and then searches data brokerage sites for their personal data, including phone numbers and addresses. This proof-of-concept highlights significant privacy risks, although the students have stated they won’t release the source code publicly.
Conclusion: This week’s AI news highlights both exciting advancements and potential risks. From substantial funding rounds and new open-source models to expanded language support and privacy concerns, the AI landscape continues to evolve rapidly, presenting both opportunities and challenges for the future.
