ai Major AI Developers Secretly Train Models on YouTube Transcripts An investigation reveals that tech giants like Apple and Nvidia are using transcribed YouTube videos to train AI models without creators' permission. This practice directly violates YouTube's own platform rules.
google deepmind Google DeepMind AI Solves Olympiad Math at Silver-Medal Level Two Google DeepMind AI programs successfully solve four out of six problems at the 2024 International Mathematical Olympiad, achieving a score equivalent to a human silver medalist.
meta Meta Unveils Llama 3.1, a Frontier-Level Open Source AI Model Meta releases Llama 3.1, featuring a massive 405B model that rivals top closed-source systems. The updated suite brings longer context windows, multilingual support, and new tools for building custom AI agents.
google Google Launches Gemini 1.5 Pro Update With Dedicated App and Live Chat Google introduces a major Gemini 1.5 Pro update featuring a dedicated app and a live conversation feature to rival ChatGPT-4o. The upgrade offers an industry-leading context window capable of processing massive documents and long videos.
youtube YouTube Allows Users to Request Removal of AI Deepfakes YouTube introduces a new privacy request process that lets individuals ask for the removal of AI-generated deepfakes that mimic their face or voice. The updated policy shifts the platform's approach from merely labeling synthetic media to actively taking it down upon complaint.
alphafold 3 AlphaFold 3 Ushers in New Era for AI-Driven Drug Discovery A recent review highlights Google DeepMind's AlphaFold 3 as a transformative leap in artificial intelligence for biomolecular structure prediction, offering unprecedented capabilities for drug design and therapeutic development.
ai Harvard Study Finds AI Models Recognize Moral Complexity But Ignore It Researchers at Harvard Kennedy School's Allen Lab discover that leading AI models mimic human moral concern but ultimately rely on a hidden value hierarchy when making decisions.
openai OpenAI Introduces CriticGPT to Catch AI Hallucinations OpenAI unveils CriticGPT, a new model designed to review and find errors in code generated by ChatGPT. This AI critic aims to solve the growing problem of hidden hallucinations as language models become more advanced.
apple intelligence Apple Unveils New Machine Learning and AI Tools for Developers at WWDC24 Apple introduces Apple Intelligence and enhanced machine learning frameworks that allow developers to build intuitive, privacy-focused AI features directly into their apps. The new WWDC24 guide provides comprehensive sessions and documentation to help creators integrate these advanced capabilities.
nvidia Nvidia Overtakes Microsoft to Become World's Most Valuable Company Nvidia surpasses Microsoft to claim the top spot as the most valuable public company globally, driven by massive demand for its AI chips. The chipmaker's market valuation reaches $3.33 trillion amid an unprecedented surge in artificial intelligence spending.
nvidia NVIDIA Overtakes Microsoft to Claim Title of World's Most Valuable Company NVIDIA reaches a staggering $3.335 trillion market capitalization, officially surpassing Microsoft to become the world's largest public company. This historic milestone is driven by explosive demand for the company's AI hardware and data center divisions.
microsoft Microsoft Delays Recall Feature After Intense Privacy Backlash Microsoft scales back the launch of its AI-powered Recall feature, restricting it to Windows Insider testers following severe security and privacy criticisms. The tool now requires extra authentication and switches off by default.
microsoft Microsoft Recall Delay Exposes Deep Windows User Privacy Fears Microsoft delays its AI-powered Recall feature after security researchers reveal critical vulnerabilities that expose user screenshots. The controversy highlights growing tension between rapid AI integration and user privacy in Windows 11.
apple Apple Unveils Private Cloud Compute for Secure Cloud AI Processing Apple introduces Private Cloud Compute, a new system that brings robust security and privacy to cloud-based AI tasks. This innovative architecture ensures that not even Apple can access a user's personal data during complex processing.
apple Evaluating Apple's Private Cloud Compute Security for Enterprise Use Apple's Private Cloud Compute aims to bring generative AI to enterprise devices without compromising user privacy. However, business environments require verifiable proof and compliance trails that go beyond Apple's hardware assurances.