Meta AI unveils WavFlow, a groundbreaking multimodal framework that generates synchronized, high-fidelity audio directly from video and text inputs, matching top latent-based models on major benchmarks while making its codebase publicly available on GitHub.
Andrej Karpathy's free 'Neural Networks: Zero to Hero' course on GitHub is empowering learners worldwide to build GPT models from scratch through 8 hands-on video lectures covering everything from backpropagation to Transformer architecture and Byte Pair Encoding tokenization.
Spotify engineers reveal a powerful two-stage AI development strategy, combining LLM evaluations and A/B testing in a feedback-driven funnel that filters weak candidates early and validates real-world impact, making both tools smarter over time.
ByteDance launches Lance, a compact 3B-parameter multimodal AI model that handles image and video understanding, generation, and editing in a single unified framework, delivering competitive benchmark performance against larger models while remaining fully open-source on Hugging Face.
Alibaba unveils the Zhenwu M890 AI chip with triple the performance of its predecessor, announces a multi-year chip roadmap, and launches a new AI model and server system, backed by a massive 380 billion yuan infrastructure investment as China races to reduce reliance on U.S. chips.
Google declares the 'agentic Gemini era' at I/O 2026, revealing 900 million monthly users, a new Gemini 3.5 Flash model four times faster than competitors, and a sweeping lineup of autonomous AI products including Ask YouTube, voice-powered Docs Live, and AI eyewear, backed by up to $190 billion in capital …
The Claude Code team is abandoning Markdown in favor of HTML outputs, unlocking richer, more readable documents featuring tables, SVG diagrams, interactive JavaScript elements, and responsive layouts that make technical specs and reports far more likely to actually get read.
Google launches the Gemini 3.5 model family at I/O 2026, introducing a blazing-fast Flash model four times faster than rivals at half the cost, a personal AI agent called Spark, and a suite of new tools spanning multimodal video, a redesigned app, and expanded desktop support.
Google launches Gemini 3.5 Flash, a frontier AI model four times faster than competitors, now available globally, while introducing Gemini Spark, a personal AI agent running 24/7 on users' behalf, set to enter Beta for Google AI Ultra subscribers in the US next week.
Google Search undergoes its biggest redesign in 25 years, now powered by Gemini 3.5 Flash, featuring AI agents that monitor the web, book reservations on users' behalf, and generate custom interactive apps in real time, with Personal Intelligence expanding to nearly 200 countries across 98 languages for free.