Today's stories I rated worth a take.

A six-repo workflow uses plan, implement, review, and verify loops.

The payoff is fewer wrong-but-clean changes across Go, Node, and React.

Use plan.md, tight prompts, and clean trees before bypass mode.

OpenAI, Anthropic, AWS, Azure, and Google Cloud give the US the stack China still lacks.

Cheap power helps, but cloud reach and data decide scale now.

If you are planning AI bets, watch infrastructure, workflows, and distribution first closely.

The deal covers at least 3,000 SLB-500M systems across three years.

That means a bigger, faster munitions pipeline for U.S. planners.

Watch for funding, delivery schedules, and industrial-base follow-through.

The bill would tighten DUV equipment controls and force Japan, the Netherlands to align.

Semiconductor firms and diplomats now face a sharper policy fight.

Watch the summit for tone, then Congress for the real decision.

The Busan deal and Beijing summit tilt rules toward China.

That shift raises Taiwan, export-control, and alliance risks.

Washington should stop mixing quick wins with strategy.

The Gemini laser setup reached estimated intensities near 10^23 W/cm^2.

That narrows the gap toward QED tests and vacuum physics.

Researchers now have a clearer path to larger field experiments.

Talkie-1930 still answers grammar questions without explaining its own rules.

That makes extraction the real problem, not training.

Future models could know more and stay harder to interrogate.

The universe already had thousands of images, characters, and language.

Hero shots, custom refs, and layered tools did the heavy lifting.

If you make AI video, copy the workflow, not the prompt.

Nature’s report cites conotoxins, viral design, and screening gaps.

The biggest risks sit in misuse, not computer output alone.

Readers should watch synthesis screening, access limits, and biodefense moves.

QEMU, Linux, and driver patches make passthrough work.

Gaming improves a lot; AI inference improves more.

The setup is useful today, but still very hacky.

TML-Interaction-Small uses 200ms micro-turns and hits 0.40-second latency on FD-bench.

That pushes AI from response mode into live collaboration.

Builders working on voice, vision, and agents should rethink turn-based stacks now.

Three numbers matter: 28.6%–91.4% fabricated references, 6% validated models, 2,271 syntheses.

The result is weaker correction, not just faster writing.

Use more verifiable records, logs, and review accountability now.

YellowKey bypasses BitLocker in WinRE; GreenPlasma reaches SYSTEM on Windows 11 and Server.

Attackers get easier access to protected disks and elevated shells.

Defenders should prioritize patching, hunting, and hardening Windows recovery paths now.

It handles ambiguous bytes by branching all feasible interpretations ahead of time.

That makes self-contained AArch64 binaries easier to test and sign.

Code size rises, but runtime translation overhead disappears.

CVE-2026-45185 affects versions 4.97 through 4.99.2 and can reach allocator metadata.

OpenSSL builds are unaffected, but GnuTLS servers need attention now.

Upgrade to 4.99.3 and check exposed mail servers immediately.

Two Iraqi VLCCs and Qatari LNG cargoes passed through Hormuz despite the war.

Traffic has fallen to about 5% of normal levels.

Buyers and shippers now have to plan around Iranian control.

IRGC swarms, drones, mines, and tunnels keep pressure on shipping.

That raises costs for oil and container traffic immediately.

Shippers and planners need to treat the strait as contested.

The system combines fully-duplex audio, video input, and delegated reasoning.

That makes real-time conversation more natural, but still imperfect.

The main takeaway is that scale, not frontier intelligence, is the advance.

ACP, UCP, AP2, x402, and Bedrock Agent Core split control.

That changes who gets paid and who gets blamed.

Map the stack now, because checkout power is moving upstream.

Treasury sanctions, Hormuz shipping, and FPV drone strikes keep changing fast.

The spillovers now hit oil, cloud, and diplomacy.

Readers should treat this as active operational context, not commentary.

Attackers used it to plan a zero-day mass exploit.

The exposure spans open-source agents and internet-facing deployments.

Defenders should audit skills, pin versions, and hunt now.

CVE-2026-44211 lets malicious pages read data and inject commands remotely.

That turns a coding assistant into an RCE path.

Validate origin headers, add session tokens, and avoid browsing while it runs.

The flaw let them bypass 2FA in a popular admin tool before vendors reacted quickly.

LLMs spotted the hardcoded trust assumption humans missed in that script.

Defenders should patch, hunt, and harden AI guardrails immediately across scripting pipelines now.

Applied economics can now refresh data and methods continuously.

Researchers must cite versions and supervise agents carefully.

Journals and tenure committees will have to adapt soon.