Blog

Notes from the platform layer.

Notes on AI infrastructure and platform engineering: Kubernetes, GitOps, LLM platforms, RAG pipelines, and FinOps, from building them in production. Also available as an RSS feed.

August 11, 2026

Anthropic Is Building Its Own Chips Now. Every Frontier Lab Is a Silicon Company.

Anthropic confirmed this week it is staffing a custom silicon team to co-design chips with Claude, the clearest sign yet that renting Nvidia is no longer a complete strategy.

ai-briefing
custom-silicon
chips
anthropic

August 11, 2026

Qwen3.8-Max Claims Ten Days of Autonomous Coding. The Open Weights Will Tell Us If It's Real.

Alibaba's new 2.4T flagship claims 10+ days of self-directed coding and posts near-frontier agentic benchmarks, with open weights landing this week. That last part is what actually matters.

ai-briefing
coding-agents
qwen
open-weights

August 11, 2026

OpenAI's August GPT-5.6 update ships with a 'High capability' label, and that label is the real story

OpenAI's August 6 refresh of GPT-5.6 Sol and Luna brings a reasoning slider and free-tier expansion, but the system card's High capability classification in cyber and bio is what practitioners should read carefully.

ai-briefing
openai
frontier-models
ai-safety

August 11, 2026

OpenAI's S-1 Is Weeks Away, and the Whole Industry Is About to Get Graded

OpenAI's public S-1 filing is expected within weeks ahead of a targeted September IPO, and the numbers inside it will reset how every AI company is valued.

ai-briefing
openai
ipo
ai-business

August 11, 2026

The Senate Just Moved Four Kids-and-AI Bills at Once. The Hard Part Is the Sorting Machine.

The Senate Commerce Committee advanced KOSA and three children's AI bills on August 5, and every one of them depends on knowing which users are minors.

ai-briefing
ai-regulation
kids-online-safety
us-policy

August 8, 2026

What I Look for When Reviewing an AI Platform Architecture

Six areas decide whether an AI platform survives contact with production: tenancy, egress control, model routing, observability, kill-switches, and IaC coverage.

architecture
ai-platform
security
kubernetes
review

July 30, 2026

A FinOps Loop for AI Workloads That Survives Past the First Cleanup

Most AI cost programs are one-time cleanups and the bill drifts back within a quarter. The weekly loop I run instead: attribution, anomaly detection on billing data, ranked levers, budgets as code.

finops
llm
cost-optimization
temporal
kubernetes

July 26, 2026

Open-Weight Isn't Cheap and Closed Isn't Doomed: The AI Price War by the Numbers

Kimi K3 priced itself like a frontier model, two leaderboards disagree by twenty ranks on the same product, and committed GPU capacity got more expensive while spot prices fell. What the token economics actually say about build vs buy.

ai-infrastructure
llm
finops
open-weight-models
model-selection

July 23, 2026

The Guardrail Lockout: When Your AI Provider Refuses Your Incident Response

An autonomous agent breached Hugging Face, then commercial model guardrails blocked the forensic investigation because a filter cannot tell a responder from an attacker. A second model is a continuity control, not a cost play.

ai-infrastructure
incident-response
llm
risk-management
open-weight-models

July 15, 2026

Running Temporal Workers on Kubernetes Without Dropping Activities

Temporal workers look like stateless services and are not. The deployment topology, rollout mechanics, and schedule management that survive production, plus the failures that actually page you.

temporal
kubernetes
argocd
platform-engineering