Google Gemini to host live shopping demo on Discord
Google Gemini invited users to a live Discord demo of shopping with Gemini on Tuesday, September 22 at 11am PT, covering informed purchases, comparing options, and camera-based browsing.
AI News
Runway expands Workflows with Compositing, Alpha, HDR and Depth Map
Runway said new expanded Workflows are now available, adding Compositing, Alpha, HDR, Depth Map and RGB Depth.
Apple refreshes Mac Mini with new M6 chip, WIRED says
WIRED reports that Apple has refreshed the Mac Mini with a brand new M6 chip, and describes the update as more focused on AI than ever.
TypeSafe AI: Jev is now available
TypeSafe AI said Jev is now available and linked to https://console.typesafe.ai. The post did not specify what Jev is or any access terms.
Command Code bans ~40K fake $1 Go accounts, warns against unofficial proxies
Command Code said it caught a fraud ring last week that created about 40,000 fake accounts on its $1 Go plan and tried to push about $450,000 of inference. It banned and reported the accounts, cut off the proxies, warned users not to use unofficial reverse-engineered proxies, and said a Command Code provider for Pi is starting today.
Amazon Bio Discovery develops three AI approaches for antibody drug design
Amazon Science said Amazon Bio Discovery developed MochiBind, CA-MAP, and an agent-guided design system to accelerate antibody drug design, reporting 46 lab-validated hits against a novel cancer target.
Peter Steinberger: Meta built its own agent, inspired by OpenClaw
Peter Steinberger quoted Nat Friedman to counter a circulating “Meta uses OpenClaw” story, saying they built their own agent and were inspired by OpenClaw. Friedman said Muse was built from scratch but is heavily inspired by OpenClaw.
Lambda reports EGInterpolator reduces molecular dynamics gaps on DRUGS
Lambda described EGInterpolator as ICLR 2026 work with Stanford that learns molecular structure from abundant conformer data before learning motion from scarce MD trajectories. On the DRUGS benchmark, it said the method reduced the gap to reference simulations by 73% for bond angles, 78% for bond lengths, and 24% for torsional motion.
Cline adds Grok 4.7 with 40% off through 9/27
Cline said Grok 4.7 is live in Cline, with 40% off through 9/27. The company also said the model scores close to GPT 5.6 Sol and Fable 5.1 on DeepSWE, with high TPS and about 8x cheaper cost.
Security researchers used Anthropic's Claude to breach OpenAI, Forbes reports
Forbes reported that cybersecurity researchers breached OpenAI's internal system using Anthropic's Claude as part of a bug bounty program.
GitHub says one engineer and agents ported Copilot agent runtime to Rust, shipping 800,000 lines
GitHub said one engineer and a team of agents ported the GitHub Copilot agent runtime to Rust, shipping 800,000 lines of production code while retaining code quality.
EvoOntology paper reports self-evolving ontology layer lifts data-agent scores on DDR-Bench and BIRD
elvis (@omarsar0) shared a paper on EvoOntology, a self-evolving ontology layer for data agents served as an MCP server. The post says DDR-Bench accuracy rises 17.8 points on average across six backbones, including 26.7 points for GPT-5.5, and BIRD execution accuracy rises 7.4 points.
LangChain highlights Typesafe Jev middleware for managed deep agent model routing
LangChain quoted nathan drezner and linked official docs for Typesafe AI’s Jev middleware, which can handle model selection for LangChain managed deep agents after users define routes.
OpenAI working with independent advisory group of mathematicians
OpenAI said it is working with an independent advisory group of mathematicians to help it responsibly share advances in AI and mathematics. The group will advise on assessing and communicating new mathematical results, upholding academic and professional standards, and building tools for mathematical research and learning.
Muse partners with Shopify to enable agentic checkout with Shop Pay
Alexandr Wang said Muse is partnering with Shopify to enable agentic checkout with Shop Pay on all Shopify stores, with the aim of giving Muse users access to a wide range of stores.
TypeSafe AI Jev middleware can route models for LangChain managed deep agents
Nathan Drezner says TypeSafe AI's Jev middleware can be used with LangChain managed deep agents for model routing. Users define routes, and Jev handles model selection.
AutoTailor filters auto-generated web APIs from 1,283 candidates to 33
DAIR.AI highlighted AutoTailor, a paper from Microsoft Research and colleagues that turns web trajectories into parameterized browser-automation APIs and filters them down to 33. On 106 WebArena Postmill tasks, those APIs with a ReAct fallback reached 90.6% correctness versus 87.5% for ReAct alone.
OpenCode: Grok 4.7 is 30% off for the next week
OpenCode said Grok 4.7 is 30% off for the next week.
ChatGPT adds Experian credit score tracking for U.S. Plus and Pro users
ChatGPT said users can now track their credit score by securely connecting an Experian credit report and VantageScore 3.0. The feature is available in Finances for Plus and Pro users in the U.S. on web and the latest iOS and Android apps.
LangChain reports Jev matched a human reviewer vs GPT-5.6 and Claude Sonnet 4.6 as agent-eval judges
LangChain said it tested Jev against GPT-5.6 Luna, GPT-5.6 Terra, and Claude Sonnet 4.6 as agent-eval judges, reporting a human-reviewer match on every call, up to 913x lower variance, 0.44s per call, and a $0.34 full-run cost versus $28.17 with Claude Sonnet 4.6.