News
Zhipu AI Stock Surges After Open-Sourcing GLM-5.2 Model
Zhipu AI’s Hong Kong-listed stock surged as much as 48% and closed up 32.8% after the Beijing-based firm unveiled GLM-5.2,…
News
Google DeepMind Delays Gemini 3.5 Pro to July 17 for Full Rebuild
Google DeepMind has postponed the launch of Gemini 3.5 Pro to July 17, 2026, opting to rebuild the model from…
News
SpaceX to Acquire AI Coding Startup Cursor for $60 Billion
SpaceX has agreed to acquire Anysphere, the company behind the AI coding assistant Cursor, in a $60 billion all-stock deal…
Opinions
Who Gains in an AI-Supercharged Economy?
Joseph H. Davis argues that as artificial intelligence technology matures, economic benefits will shift from AI developers to the companies…
Opinions
America’s AI Policy Is Truly Chaotic
Francis Fukuyama argues that United States AI policy has become chaotic, lacking clear institutional structures for regulation. He notes a…
Opinions
On AI’s 70th Birthday, A Warning About Excessive Regulation
Mark Jamison argues that while governments today scrutinize large technology companies through antitrust enforcement, history suggests regulators often lack foresight…
Opinions
Teens Are Turning to Chatbots for Mental Health Help. We Need Rules to Keep Them Safe
Ryan McBain argues that regulators should establish targeted guardrails for AI chatbots rather than implementing blanket bans for teen users.…
Engineering
Why Is PyTorch Compile So Fast: Kernel Fusion
PyTorch’s compiler achieves significant speedups through kernel fusion, a technique that combines dependent operations into single, efficient kernels. By fusing…
Engineering
Understanding PyTorch’s Test Infrastructure
PyTorch’s test infrastructure dynamically generates tests across devices and data types at import time, which is why CI test failures…
Engineering
Adding a GPU Without Building One
VKAE is an inference acceleration technology that optimizes GPU performance through software-level kernel improvements rather than hardware changes. The system…
Engineering
DiffusionGemma: 4x Faster Text Generation
Google released DiffusionGemma, a 26B mixture-of-experts model that generates entire blocks of text simultaneously rather than processing tokens sequentially like…
News
Nobel Laureate John Jumper Leaves Google DeepMind to Join Anthropic
John Jumper, the Nobel Prize-winning creator of AlphaFold, is departing Google DeepMind after nearly nine years to join Anthropic. His…