RuntimeWire

Real-time startup intelligence for the AI economy. Funding, model launches, infra shifts, and founder moves — the signals that matter.

Latest stories — page 18

• Devin migrated Cognition's site and recorded its own verification runs

  Jared Palmer says the Astro-to-Next.js project used recorded tests to reproduce subtle failures, though Cognition published no scope or performance data.

• Alvys puts a build-your-own AI agent layer inside its freight TMS

  Nick Darman's Foundry starts with 20-plus templates, human approval gates and a waitlist, extending the $40M Series B bet into freight operations.

• Google tests phone photos as a screen for insulin resistance

  PhotoScan estimated body-fat distribution in a clinical study, though its closest DXA comparison rests on a 132-person validation cohort.

• J-Space says its text harness pushed DeepSeek V4 Pro past Fable 5

  The reported gains are substantial, but the viral claim outruns single-run tests that mix vendor benchmark methods.

• Relay shuts down; founder Jacob Bank returns to Google as Chrome VP

  Paid users keep access through September 14th; Bank will lead Chrome's product and developer-relations teams.

• Lanyon AI discloses $10.6M round to make physics code prove itself

  Dimension led the July financing, backing a Princeton trio building AI-generated simulations with machine-checkable proofs.

• SecretSpec releases dotenv-ng, says dotenvy changed a bcrypt credential

  Domen Kozar's fork gives SecretSpec control over fixes and releases for the parser used in its credential-migration path.

• Kimi Desktop shipped Moonshot’s hidden employee gateway for off-quota Kimi, GPT and Codex access inside its public client

  Version 3.1.10 identifies the office-only service as "KTH Gateway (Internal)," requires a personal bearer token, and routes off-quota Kimi, GPT, and Codex access through a shared upstream pool.

• Anthropic's revenue run rate tops $65B ahead of its planned IPO

  The July figure is up from $9B at the end of 2025 and 38% above the pace Anthropic reported in May.

• OpenAI awards $1M to 14 projects studying AI's economic and social effects

  The six-month grants span jobs, benefits, data-center energy, biosecurity and democratic oversight, with results due in 2027.

• Sam Hogan launches Lumbridge, a RuneScape-style test world for AI agents

  The open-source experiment lets Claude, Codex, Hermes and other coding agents control persistent characters through a TypeScript SDK.

• Nous Research turns Hermes profiles into reusable specialist Bots

  The desktop update gives each named Bot its own model, memory and skills, with cross-Bot communication inside the same app.

• Khosla and Emanuel argue autonomous AI could outperform doctors using AI

  The JAMA Perspective draws on simulated studies and challenges medicine's prevailing physician-led deployment model.

• Reddit tests audio and video versions of viral text posts

  The mobile experiment adds a "play" option while preserving the original post, comments and community discussion.

• Amazon buys and destroys printed books for AI training data

  A tracker hidden in a roughly 1,000-book order led 404 Media to a Las Vegas site where workers cut bindings and scan pages.

• Browser Use launches macOS Harness for direct control of Mac apps

  The open-source Python layer gives coding agents access to native apps, logged-in browsers, files and shell commands through six primitives.

• Sonic Fire Tech raises $15M to push acoustic suppression through fire codes

  Former NASA engineer Geoff Bruder plans to spend the round on hiring and the tests needed to move infrasound from demos into regulated buildings.

• Terra adds $18M, closing a $52M seed for defense manufacturing

  Nathan Nwachuku and Maxwell Maduka plan to keep production in Africa while opening London access to defense buyers, capital and talent.

• 3M expert told ChatGPT to build zero-fault case before $61M explosion verdict

  Discovery produced 350 pages of prompts, including a request to show 3M was "0% at fault," before jurors assigned 30% of the blame to 3M.

• MiniMax pitches H3 as a video model for consistent game-sprite motion

  MiniMax is pitching its general-purpose multimodal H3 model for game-sprite animation, though the supplied materials include no independent comparative benchmark.

• CarSignal launches a $200-a-month AI operating system for auto repair shops

  The YC-backed startup says more than 40 shops use its software for booking, diagnostics, estimates, parts and payments.

• World Liberty links its tokens to Chinese AI models under US scrutiny

  WorldClaw takes USD1 payments and offers 43 models tied to Chinese companies flagged by the Trump administration, Reuters found.

• Uber backs Zipline to target 1 million daily drone deliveries by 2029

  The Uber Eats integration will begin in Houston and Dallas before expanding across dozens of U.S. cities.

• Wiz says Red Agent exploited a Snowflake workflow flaw introduced by Copilot

  A June workflow change let a crafted issue title execute commands and expose a Jira token for five days before Snowflake patched it.