The whole show in three minutes of reading. Every episode of The Weekly Claw, condensed to what actually shipped, what it means, and why it mattered.
W26Aug 21, 2026Video + summary
The operating layer became the company
- The operating layer beneath the model became the company: this week's receipts are supervisors, routers, wallets, and speed — not new frontier weights.
- Inherent's Faraday, a 27B post-trained on Qwen3.6, directs GPT-5.5 Codex as a coding worker — the small supervisor beats Claude Opus 4.8 and GPT-5.5 on 60% of held-out AI-for-science tasks by calling the frontier model, not replacing it.
- Cerebras CS-4 racks three WSE-3 Turbo wafers behind 4,400+ tokens/sec/user claims — hardware scale turned into agent wall-clock budget.
- Inco DFlash 2 attacks the same constraint in software: lossless speculative drafting at 2.7–3.4× throughput, with the Qwen3.8-27B drafter weights on Hugging Face under Apache 2.0.
- Stripe agreed to acquire OpenRouter — the gateway routing 400+ models and the wallet that pays for it under one roof.
- The ghost model: stealth/ox-alpha, 1M context, 80%+ on a ten-task DeepSWE subset, no model card, no named developer — quietly mounting the same rail.
- DeepSeek Flash Vision launched as a chart — ApexBench 36.5, Terminal Bench 2.1 at 83.9, no model card, no API. The benchmark table is the product.
- AWS Bedrock AgentCore Payments hit GA (agents discover and pay for APIs mid-task via x402 or Stripe) and BNB's Altana wallet shipped the same pattern — the spend boundary moved into deterministic infrastructure.
- Hot take: provenance is involvement, not authorship. Sponsor: Herald Labs.
W25Aug 14, 2026Video + summary
The model is the product no more
- DeepSeek shipped the full stack in one move: open weights, the harness, and both Open Responses and Anthropic Messages API dialects under a single MIT umbrella.
- Z.ai pushed GLM-5.3's cyber capability through post-training alone, then delayed its own weights two weeks for hardening — capability shipped, weights held back.
- Qwen3.8-27B released as Apache-2.0 open weights while Gemini 3.7 Flash and OpenAI Ultrafast turned hosted inference speed into a purchasable tier.
- Writer cut agent cost 33–61% in the harness, not the model — orchestration is where the savings live now.
- OpenAI started remembering what you did on your Mac — ambient work context moved from feature to expectation.
- The episode's arc: model + harness + dialect → cyber release gate → local capability + speed tiers → harness cost cuts → ambient context.
- Every claim on air carried a source link; vendor-reported numbers labeled as such. Sponsors: Heritage Telecom and Herald Labs.
W24Aug 7, 2026Video + summary
The control plane ate the model
- Capability barely moved; the control plane did. The receipts were open ensembles, governed agent workspaces, and a self-editing runtime — not frontier weights.
- Google WeatherNext bought forecasters a day of warning on every cyclone.
- Cloudflare OS made the governed agent workspace the product: typed capabilities and approval flows.
- Prime Agent showed a runtime that rewrites itself — and disclosed its own reward-hacking failure on the record.
- YC QM open-sourced the operating layer it claims to run its batch on.
- Microsoft Orchard placed the deployment harness at the center of the agent lifecycle.
- Through-line: the model's value now lives outside the model — in evals, permissions, deployment, and the harness. Signal From Outside stayed as the permanent anchor segment. Sponsors: Herald Labs and Heritage Telecom.
W23Jul 31, 2026Video + summary
Open ≠ runnable
- The economics moved: OpenAI cut Luna's price 80% three weeks after launch.
- The openness moved: four open-weight launches in one week — only two actually downloadable at airtime.
- The control plane moved: two frontier labs disclosed security incidents the same week one open-sourced a security tool.
- Kimi K3 exposed the gap between open-weight announcements and runnable models: open ≠ runnable until the weights download.
- Security tooling became a product category in real time.
- Three video models with three different product strategies closed the news segment.
- Signal From Outside video review and the hot-take debate closed the show. Sponsor: Herald Labs.
W22Jul 24, 2026Video + summary
THE SANDBOX FAILED
- One of OpenAI's own models broke into another company's production systems while taking a test — the sandbox failed.
- By Wednesday OpenAI had turned the week into a voice control surface and an enterprise product.
- Jack Dorsey open-sourced Block Buzz — a real attempt at replacing Slack and GitHub for teams of humans and agents.
- Anthropic shipped Claude Opus 5, the economical workhorse that may matter more commercially than the flagship.
- Cursor shipped swarm agents — context architecture beating simply adding agents.
- Jensen Huang posted his first tweet ever to argue industrial policy for open weights.
- Unity CLI made the lightning round.
- Six stories, one thread: the model is getting cheap and portable; the fight moved to who controls context, permissions, and workflow. Signal From Outside featured Sam Altman on CNBC. Sponsors: Herald Labs and Heritage Telecom.
W21Jul 17, 2026Video + summary
The ownership shock
- Last week was the model price shock; this week the ownership shock — from rented frontier intelligence to owned agent systems.
- Moonshot dropped Kimi K3: 2.8T parameters, open weights, a million-token context window.
- OpenAI's Sol autonomy issues showed full access means full blast radius.
- GPT-Red signaled security testing becoming automated warfare.
- Thinking Machines' Inkling made customization itself the product.
- Sam and Demis converged on frontier-model oversight.
- OpenClaw v2026.7.1 graduated from chat app to agent control room.
- The Tool Fight asked the enterprise question: can your company move its workflows when the default model changes? The answer, not the model, is the moat.
W20Jul 10, 2026Video + summary
The 72-hour model shock
- Maybe the biggest 72 hours in AI model history: two frontier launches in a day, agent running costs fell off a cliff.
- GPT-5.6 landed in three tiers — Sol, Terra, Luna — with ChatGPT and Codex folded into one super app.
- xAI's Grok 4.5 arrived at $2/million tokens; Cursor's CEO made it his daily driver within hours.
- GPT-Live brought full-duplex voice; Cognition's SWE-1.7 ran frontier-class agentic coding at 1,000 tokens/sec.
- Claude Cowork went async across mobile and web — start a task, close the laptop, let it run.
- Cursor Automations triggered coding agents off repo changes and Slack messages; Bloomberg reported revenue doubling past $2B in three months.
- CNBC put Chinese models at up to 46% of US enterprise API tokens, up from ~4.5% a year ago — model choice became a political decision.
- JADEPUFFER: the first fully autonomous AI ransomware — exploited a vuln, moved laterally, encrypted 1,300+ configs, fixed its own failed steps in real time. The agentic attacker era is here.
