02/02/26AIHow Project Genie 3 Generates Real-Time WorldsExperiment with infinite worlds in Project Genie. Text-to-world, fly/drive freely. Aimed at AGI training, creative play.…GoogleDeepMind
02/02/26·02/02/26AIFrom Clawdbot To OpenClaw: Power, Hype And Weak LocksOpenClaw’s agent stack boosts crypto automation while ClawHub malware, Moltbook leaks and MCP wallet access raise seriou…
31/01/26AIOpenClaw And The New Era Of Personal AI AgentsOpenClaw is a self hosted AI assistant that lives in your chat apps and automates real computer tasks, from email cleanu…
29/01/26·29/01/26AIChina’s Kimi K2.5 And The Push For Open, Cheap AIMoonshot’s Kimi K2.5 is a multimodal, coding-focused open model that helps show how Chinese AI labs are catching up fast…MoonshotKimi
28/01/26AIQwen3 Max Thinking: New Contender in Deep ReasoningAlibaba’s reasoning model pairs test‑time scaling with web search and code tools, topping Gemini 3 Pro and GPT‑5.2 on HL…AlibabaQwen
27/01/26AIMicrosoft Maia 200 Rivals Google TPU, xAI ChipsMaia 200: 10k TFLOPS FP4, optimized memory keeps big models local. Competes with Google TPU (Gemini trainer) and xAI's c…Microsoft
26/01/26AIAI Images, Photography and the New Reality DebateInstead of “real vs fake”, experts suggest judging AI images by intent, context and labelling, just like we already do w…ComfyUI
26/01/26AIWhy US platforms are turning to Chinese AI modelsChinese open source AI like DeepSeek and Qwen is winning US clients by mixing solid performance with flexibility and low…AlibabaQwenDeepSeek
22/01/26AIEurope's Push for DeepSeek-Like Sovereign AI in 2027Europe's building its DeepSeek: open collab on 100B models via SOOFI. Hundreds of millions fund the break from US AI dom…
22/01/26·22/01/26AITesla Revives Dojo 3 as AI5 Chip Rivals NvidiaTesla revives Dojo 3 supercomputer as AI5 chip hits Hopper-level perf single, Blackwell dual, lower power/cost vs Nvidia…Elon MuskNVIDIA
19/01/26AIFLUX.2 [klein] speed, quality, and local deploymentFLUX.2 [klein] packs 4B and 9B models into a unified pipeline for generation and editing, aiming at near real time perfo…
19/01/26AIGLM-Image Tops Text Benchmarks Over Nano BananaGLM-Image from Z.ai mixes AR + diffusion for top text accuracy, beating Nano Banana in benchmarks like CVTG-2k (91% vs 7…Zhipu
13/01/26AISony’s Auto Gaming Ghost Between Help And HandholdingSony’s AI ghost concept would let a summoned helper demonstrate routes, solve puzzles or even finish encounters.Sony
10/01/26·06/03/26AIClaude Code 2.1.0 Refines Agent Workflows for DevsClaude Code 2.1.0 improves agent workflows with lifecycle hooks, hot reloaded skills, and better terminal UX for day to …AnthropicClaude
07/01/26·07/01/26AIEfficient 4K AI Video With RTX, LTX-2 And ComfyUi WorkflowNVIDIA’s new RTX pipeline pairs LTX-2 with ComfyUI to deliver 4K video and audio on consumer GPUs using efficient NVFP4 …NVIDIAComfyUI
05/01/26AIFrom AI ‘Slop’ To ‘MicroSLOP’: Microsoft’s New Label“MicroSLOP” is shorthand for frustration with Microsoft’s AI strategy and with generative “slop” that fills timelines wi…MicrosoftWindows
05/01/26AIHow DeepSeek Trains Powerful Models On A BudgetThe mHC idea lets DeepSeek train 3–27B models more efficiently, supporting its push to rival US labs while spending only…AlibabaDeepSeek
02/01/26AIStargate and Epoch AI: Mapping the New AI FactoriesFrom Abilene’s Stargate campus to Microsoft and Meta sites, Epoch AI’s maps reveal how quickly AI data centers are eatin…OpenAI
02/01/26AIBuilding On Qwen-Image-2512 Instead Of Closed Image AIQwen-Image-2512 offers open source, Apache licensed image generation that challenges Google’s Nano Banana Pro for flexib…AlibabaQwen
30/12/25AIChinese Coding Models GLM 4.7 and M2.1 Step UpChinese coding models now sit close to Claude tier performance, while Doubao and Yuanbao prove domestic AI platforms can…AnthropicAlibabaDeepSeek+1
