Latest dispatches

Every published story, ordered for readers, crawlers, and AI citation tools alike.

A chronological archive built for scanning rather than guessing, with visible bylines, stable article routes, and fast jumps back into category archives.

Chronological archiveStable article URLs

Stories: 105
Bylines: 4
Categories: 6
Latest story: Apr 11, 2026

AI Products/Apr 11, 2026/Updated Apr 11, 2026/11 min read

Seedance 2.0 takes Sora's abandoned AI video crown

ByteDance slipped Seedance 2.0 into the US on April 7, just after Sora shut down. That timing was not accidental, and the copyright fight did not disappear on the way in.

Maya HalbergStaff Writer

Editorial illustration of a vacant AI video throne being claimed by ByteDance creator apps, with CapCut, Dreamina, and Pippit visual cues in the foreground and Hollywood legal notices gathering at the edges.

AI Policy/Apr 10, 2026/Updated Apr 11, 2026/10 min read

Perplexity's Incognito Mode didn't hide anything

A class-action suit alleges Perplexity piped millions of AI chat transcripts to Meta and Google through hidden trackers. Incognito Mode did nothing to stop it.

Lena OrtizStaff Writer

Magnifying glass cracking a smartphone screen to reveal surveillance feeds flowing from an AI chat interface into hidden ad-network pipes, in noir cinematic style with deep teal and amber tones.

AI Products/Apr 10, 2026/Updated Apr 11, 2026/10 min read

Claude Cowork gets the enterprise control layer

Anthropic's Apr. 9 update gives Claude Cowork the RBAC, spend caps, analytics, and telemetry that turn a flashy desktop agent into something companies can actually roll out.

Idris ValeStaff Writer

Editorial illustration of an enterprise Claude Cowork rollout dashboard layered over a desktop agent workspace, with team permissions, spend controls, and workflow telemetry visible as the main visual idea.

AI Infrastructure/Apr 10, 2026/Updated Apr 11, 2026/11 min read

AWS Agent Registry turns sprawl into a control layer

AWS Agent Registry is a preview governance layer that catalogs agents, tools, and MCP servers across environments so enterprises can approve, reuse, and control sprawl.

Lena OrtizStaff Writer

Editorial illustration of an enterprise agent registry control layer showing governed records for agents, tools, and MCP servers flowing across AWS, other clouds, and on-prem systems.

AI products/Apr 10, 2026/Updated Apr 11, 2026/10 min read

Gemini notebooks pull NotebookLM into the app

Google's new Gemini notebooks sync with NotebookLM and turn chats, files, and instructions into a shared workspace built for longer-running work.

Maya HalbergStaff Writer

Editorial illustration of a shared Gemini notebook workspace syncing chat threads, source cards, and uploaded files with a NotebookLM research panel.

AI Policy/Apr 10, 2026/Updated Apr 11, 2026/10 min read

OpenAI teen safety blueprint becomes ChatGPT's U18 mode

OpenAI is turning its teen safety blueprint into product behavior, with a separate U18 ruleset, age prediction defaults, tighter guardrails, and parental controls.

Lena OrtizStaff Writer

Editorial illustration of a split ChatGPT-style interface where an adult lane stays open while an under-18 lane is surrounded by tighter safety rails, age checks, and parent-control surfaces.

AI Products/Apr 9, 2026/Updated Apr 11, 2026/11 min read

OpenAI's enterprise AI stack gets specific

OpenAI is packaging Frontier, ChatGPT Enterprise, Codex, and AWS runtime as one enterprise stack, with CyberAgent as the clearest proof point yet.

Idris ValeStaff Writer

Editorial illustration of an enterprise AI control floor where a central OpenAI operating layer routes context, permissions, and work between ChatGPT Enterprise, Codex, and AWS-backed runtime lanes.

AI Infrastructure/Apr 9, 2026/Updated Apr 11, 2026/7 min read

Meta's $21B CoreWeave deal runs through 2032

CoreWeave's filing-backed Meta agreement runs into the Vera Rubin era, a reminder that custom silicon has not made Meta independent of giant external AI cloud buys.

Maya HalbergStaff Writer

Editorial illustration of Meta reserving a long-horizon external AI compute lane from CoreWeave, with future-generation Rubin hardware and data-center capacity extending deep into the next decade.

Open Source AI/Apr 9, 2026/Updated Apr 11, 2026/11 min read

Safetensors joins PyTorch Foundation weight stack

Hugging Face has moved Safetensors into the PyTorch Foundation, turning the default open-model weight format into vendor-neutral infrastructure for the wider stack.

Idris ValeStaff Writer

Editorial illustration of Safetensors moving from Hugging Face into the PyTorch Foundation, with model-weight blocks, governance cues, and open-source infrastructure in one frame.

AI Infrastructure/Apr 9, 2026/Updated Apr 11, 2026/12 min read

TorchTPU gives PyTorch a straighter path to TPU pods

Google's new TorchTPU stack gives PyTorch teams a more native route into TPU training and serving through eager execution, torch.compile, and MPMD-aware distributed support.

Lena OrtizStaff Writer

Editorial illustration of PyTorch workloads crossing a cleaner runtime bridge into a Google TPU pod, with compiler layers and distributed lanes visible.

AI Products/Apr 9, 2026/Updated Apr 11, 2026/12 min read

Meta Muse Spark is really a distribution launch

Meta is using Muse Spark to put a proprietary reasoning model in Meta AI now, then roll it into WhatsApp, Instagram, Facebook, Messenger, AI glasses, and an API preview.

Maya HalbergStaff Writer

Editorial illustration of Muse Spark as one proprietary AI core routed into Meta AI, WhatsApp, Instagram, Facebook, Messenger, and AI glasses.

Ai Tools/Apr 8, 2026/Updated Apr 11, 2026/10 min read

GLM-5.1 coding agent and the 8-hour shift

Z.AI pitches GLM-5.1 as an 8-hour coding agent and documents manual routes into Claude Code, OpenClaw, and OpenRouter. Access is real. Proof is pending.

Idris ValeStaff Writer

Editorial illustration of a long coding shift routed through GLM-5.1, with a steady agent loop spanning planning, edits, tests, tool calls, and a visible handoff into Claude Code, OpenClaw, and OpenRouter-style lanes.