September 24, 2026
Quick hits from across the AI world in the last 24 hours.
- ⚡ Quick hits
- AIGC Meshy showcased a memorial space created with Meshy and ChatGPT, allowing visitors to leave flowers and messages for a pet. X / Meshy ↗
- Industry Sakana AI won Japan’s 2026 Startup Award in the communications category and presented an AI defense command-and-control case. X / Sakana AI ↗
- Agents DHH, Ruby on Rails’ creator, says AI tools have ended his habit of writing code by hand. The Decoder ↗
- Funding Lightspeed is targeting a $250 million India fund focused on early-stage artificial-intelligence startups. TechCrunch ↗
- Open Source Gradio showed an ML intern training a LoRA in HuggingChat from one prompt, including data rendering, baselines, and evaluation. X / Gradio ↗
- AIGC Gradio demonstrated an image tool that turns a transparent PNG into new object views in about 25 seconds. X / Gradio ↗
- Open Source Fastino released GLiNER2.5-Decide, a 340-million-parameter open-weight decision model designed to run on CPUs. MarkTechPost ↗
- Agents Arena reported that Grok 4.7 entered Agent Arena at number sixteen with a net improvement score of 3.96 percent. X / Arena ↗
- AIGC Runway’s WorldPrompt explores engineering methods for real-time interactive worlds built with generative AI. Latent Space ↗
- Agents Google Cloud explained agent harnesses as the layer combining an LLM with tools, context, and execution controls. X / Google Cloud ↗
- AIGC Tripo began a semester-long collaboration with Carnegie Mellon’s Game Creation Society to explore AI-native game development. X / Tripo ↗
- AIGC Pika connected Grok Bots to its API, enabling image, video, and audio generation across more than 120 models. X / Pika ↗
- Research Artificial Analysis launched Terminal-Bench-Science, where GPT-6 Astra and Claude Opus 5.5 led rivals by roughly twenty points. X / Artificial Analysis ↗
- Industry Databricks showed how governed data can serve Jev-style decision models through serverless GPUs and an AI functions workflow. X / Databricks ↗
- Agents Baseten introduced a workflow that turns LangSmith traces into datasets and starts model post-training through Baseten Loops. X / Baseten ↗
- Industry Qualcomm is promoting one AI software stack across device categories and plans native Windows 11 support for Mojo. IT之家 ↗
- Safety Proofpoint’s latest security research argues that AI-agent defenses must understand user intent as agents enter the workforce. SiliconANGLE ↗
- Agents Neo4j is positioning knowledge graphs as shared context that helps AI agents reason across enterprise data and workflows. SiliconANGLE ↗
- Models PrismML demonstrated a one-bit, two-billion-parameter Bonsai model designed to run locally on Qualcomm smart glasses. TechCrunch ↗
- Agents Perplexity says its Photon retrieval system now handles search and ranking, cutting p99 latency from about 800 milliseconds to 65. X / Perplexity AI ↗
- AIGC Meta introduced a phone-based AI game creation experience that lets users build games directly for Horizon devices. The Verge ↗
- Agents WSO2 launched Agent Manager to help enterprises govern the growing number of AI agents across their organizations. InfoQ 中文 ↗
- AIGC Runway added Draft mode for Seedance 2.5, generating exploratory videos faster and with fewer credits before final rendering. X / Runway ↗
- Open Source Black Forest Labs launched FLUX 3 Action, an open robotics model aimed at linking visual understanding with physical actions. The Decoder ↗
- AIGC Google Photos released an AI virtual closet inspired by Clueless for Android and iOS users. TechCrunch ↗
- Agents GitHub showed an agent automation that performs first-pass review on batches of Dependabot pull requests. X / GitHub ↗
- Agents AWS published a reference architecture using AgentCore Gateway and MCP to connect agents securely across multiple AWS accounts. AWS ↗
- Industry Sakana AI appointed deep-learning pioneer Jürgen Schmidhuber as chief scientific advisor. X / Sakana AI ↗
- Open Source Researchers reported nearly sevenfold higher DeepSeek inference throughput on PCIe GPUs after kernel and communication redesigns. 量子位 ↗
- Models Liquid AI introduced LFM2.5-VL-DSpark, a draft-model approach intended to accelerate vision-language inference. Hugging Face ↗