AI
Artificial intelligence: LLMs, agents, diffusion, vision, NLP and the latest from top labs.
593 published articles
AI Labs & Research
MiniMax's M3 just wrote its own CUDA kernel, and opened the code
MiniMax M3 scores 83.5 on BrowseComp, edges past Opus 4.7, and handles up to 1M tokens natively. In a remarkable autonomy test, it self-optimized a GPU kernel from 7.6% to 71.3% peak utilization without human intervention.
2026-07-09
Voice AI race
OpenAI's GPT-Live is racing to make voice chats with AI feel real
OpenAI releases GPT-Live, a new voice model for ChatGPT that aims to make spoken interactions with the AI more fluid and human-like. The update is rolling out now, but key details are still under wraps.
2026-07-09
DeepSeek's DSpark framework rearms speculative decoding for high-concurrency serving
Semi-autoregressive decoding just broke the 85% speed barrier in production AI inference
A new speculative decoding framework from DeepSeek tackles the two bottlenecks that have limited parallel drafters: suffix decay and wasteful verification. DSpark achieves 60, 85% faster generation speeds in production by coupling a semi-autoregressive architecture with a confidence-scheduled scheduler that prunes low-value tokens before the target model verifies them.
2026-07-08
NLP history
The benchmark that made language models speak: how 2018's glue bet changed ai forever
The GLUE benchmark, launched in 2018, transformed natural language processing by providing a standardized yardstick for language understanding. Its legacy lives on in every modern LLM benchmark, from SuperGLUE to the latest arena-style evaluations that define today's AI race.
2026-07-08
Frameworks & Tools
Microsoft's Flint hides the chart boilerplate so AI agents stop drawing wrong axes
Microsoft Research introduced Flint, a visualization intermediate language that helps LLMs and AI agents create polished charts without hand-coding low-level parameters like scales and axis formatting. In a study across three models, Flint outperformed direct Vega-Lite generation, and is already used internally in Data Formulator.
2026-07-08
AI Safety
Anthropic's jailbreak severity scale is a proposal that could reshape ai safety regulation
Anthropic proposes a four-axis scoring system for AI jailbreak severity, from 'informational' to 'critical,' and details the classifiers that block dangerous cybersecurity uses of Fable 5. The framework, developed with Glasswing partners, aims to standardize how the industry and regulators talk about model misuse.
2026-07-08
Microsoft Build 2025
Microsoft's new platform gives scientists a governed factory for AI agents
Microsoft Discovery is now generally available, offering organizations a governed platform for agentic AI in science and engineering. The move signals Microsoft's push to capture enterprise R&D workflows with built-in compliance and orchestration.
2026-07-08
Cloud Infrastructure
How alibaba cloud pushed its way into 20 gartner quadrants and what it means for the ai cloud race
Alibaba Cloud's analyst report haul for 2025 and 2026 shows a company intent on outflanking AWS and Azure in Asia and Oceania. With four Gartner emerging leader positions in generative AI, a leadership ranking in enterprise MaaS, and a first-ever inclusion in AI coding agents, the Chinese cloud giant is making a case for itself as a viable alternative, if trust and latency concerns can be overcome.
2026-07-08
AI Model Release
MiniMax launches M2.7 model with strong software engineering and office productivity skills
MiniMax's M2.7 model delivers strong results in software engineering benchmarks and professional office tasks, with a 97% skill adherence rate on complex instructions and an ELO score of 1495 on GDPval-AA.
2026-07-08
Product Expansion
Anthropic just bet on scientists, teams, and Seoul. The old chatbot playbook is dead.
Anthropic released Claude Sonnet 5, a frontier AI model for coding and professional work, along with Claude Science for researchers and Claude Tag for team collaboration. The company also opened a Seoul office and published policy proposals. The old chatbot playbook is dead.
2026-07-08
Product Launch
Anthropic Launches Claude Science, an AI Workbench for Scientific Research
Claude Science by Anthropic brings together databases, analysis tools, and compute management into one AI-powered workbench. Researchers can analyze literature, run multi-step analyses, and generate publication-ready figures with auditable histories.
2026-07-08
Usage Policy
Anthropic Updates Usage Policy: New Rules for High-Risk AI Applications
Anthropic has released an updated Usage Policy categorizing permissible and prohibited uses of its AI models, with stricter requirements for high-risk sectors such as healthcare, finance, and legal. The policy applies universally to all users and introduces mandatory human oversight and disclosure for consumer-facing decisions.
2026-07-08