News
Mozilla's agentic AI pipeline turns Claude Mythos Preview loose and finds 271 unknown Firefox vulnerabilities
55+ min ago (558+ words) the-decoder. com Mozilla's agentic AI pipeline turns Claude Mythos Preview loose and finds 271 unknown Firefox vulnerabilities - Mozilla developers used Claude Mythos Preview to identify and fix 271 unknown security vulnerabilities in Firefox 150, contributing to a record 423 resolved security issues in April....
Open AI's new voice model brings GPT-5-level reasoning to real-time conversations
15+ hour, 35+ min ago (487+ words) Open AI is shipping GPT-Realtime-2, GPT-Realtime-Translate, and GPT-Realtime-Whisper - a new generation of voice models built to reason, translate, and transcribe on the fly. Chat GPT has had an audio mode for a while, and Google offers a similar real-time conversation…...
Europe's answer to AI regulation complexity is to just delay most of it
16+ hour, 42+ min ago (329+ words) the-decoder. com Europe's answer to AI regulation complexity is to just delay most of it The European Commission, the European Parliament, and the Council of the EU have agreed on simpler rules for artificial intelligence. The deal is based on…...
The US and China are considering formal talks on AI
23+ hour, 6+ min ago (187+ words) the-decoder. com The US and China are considering formal talks on AI The US and China are exploring official talks on artificial intelligence, the Wall Street Journal reports. The topic could come up during the summit between Donald Trump and…...
AI models follow their values better when they first learn why those values matter
21+ hour, 34+ min ago (374+ words) A study from the Anthropic Fellows Program shows that training a language model on texts explaining its intended values before teaching it specific behaviors leads to significantly better adherence to those values, even in situations never encountered during training. A…...
Claude's new "Dreaming" feature is designed to let AI agents learn from their mistakes
23+ hour, 15+ min ago (251+ words) Anthropic is expanding its Claude Managed Agents platform, launched in April, with three new features for making AI agents more reliable and adaptive in real-world use. A novel addition is "Dreaming," a process that reviews past agent sessions, spots patterns,…...
Google speeds up Gemma 4 threefold with multi-token prediction
1+ day, 18+ hour ago (205+ words) the-decoder. com Google speeds up Gemma 4 threefold with multi-token prediction Google has released multi-token prediction drafters (MTP) for its open AI model family Gemma 4, designed to speed up text generation by up to three times. LLMs normally generate text one…...
Open AI built a networking protocol with AMD, Broadcom, Intel, Microsoft, and NVIDIA to fix AI supercomputer bottlenecks
1+ day, 15+ hour ago (187+ words) Open AI has teamed up with AMD, Broadcom, Intel, Microsoft, and NVIDIA to develop a new network protocol called MRC, short for Multipath Reliable Connection. MRC is designed to make data transfers between GPUs in large AI supercomputers faster, more…...
Google and Meta race to build personal AI agents as Anthropic and Open AI pull further ahead
1+ day, 21+ hour ago (423+ words) Google and Meta are both testing personal AI agents designed to handle everyday tasks on their own. The push is a direct response to the lead built by Anthropic, Open AI, and Microsoft. Google has shut down a previous agent…...
Anthropic co-founder maps out how recursive AI improvement could outpace the humans meant to supervise it
2+ day, 22+ hour ago (312+ words) Jack Clark argues in a long essay that the building blocks for AI systems training their own successors are largely in place. He puts the odds at 60 percent by the end of 2028. In his newsletter Import AI, Anthropic co-founder Jack…...