Writing

Notes from building AI-era systems that have to hold up.

Insights on the Laravel AI SDK, AI agents, executive leadership, technical excellence, and software development.

Oct 15, 2026

10 min

Technical excellence

What Is WebMCP? What It Means for Your Website

WebMCP lets AI agents use your website through structured tools instead of clicks. Here is what it is, how it differs from MCP, and what to do now that Shopify has switched it on for checkout.

Oct 02, 2026

10 min

Ai innovation

The Death of the Seat License: Will AI Agents End SaaS?

AI agents are breaking the link between software users and software revenue. This piece looks at what Gartner, Deloitte and others actually say about SaaS spending, which products look exposed, which defenses hold up, and what founders and buyers can do now.

Sep 23, 2026

9 min

Ai innovation

Grok Bot Roles: Own Outcomes Soft Pyramid Operators Can Trust

Grok Bot works for Soft Pyramid operators when each Bot owns a repeatable outcome, not a pile of open questions. Start with read-and-prepare work, review the result, then add approved actions. Eight starter roles show how.

Sep 19, 2026

7 min

Technical excellence

n8n Production Case Study: What Survives After the Demo for US Service Firms

n8n demos look clean; production fails on empty payloads, credential drift, and silent retries. This case study walks through anonymized US service-firm patterns — enrichment, inbox triage, and meeting handoffs — and the operating rules that keep workflows alive after go-live.

Sep 17, 2026

8 min

Technical excellence

What Production AI Systems Need That Demos Don’t

Production AI systems need more than a working demo: auth, queues, rollback, and an owner at 2am. This article walks through what demos omit, what research says about pilot failure, and how Soft Pyramid judgment helps US SMBs ship systems that hold up.

Sep 14, 2026

8 min

Ai innovation

Stronger Than Us, Not Smarter: Why Superintelligence Breaks Human Prediction

Humans have built machines stronger and faster than people — but not more generally intelligent. This article examines why that gap matters for prediction and control, what lab evidence actually shows, and how SME operators should set autonomy budgets without hype or panic.

Sep 12, 2026

9 min

Ai innovation

Pacing the Frontier Explained: What Amodei’s Slowdown Proposal Means for Operators

Anthropic CEO Dario Amodei’s September 2026 essay argues frontier labs should pace capabilities so safety work can keep up, citing recursive self-improvement and the OpenAI-Hugging Face agent incident. This article unpacks the three-step plan and what SME operators should take from it without treating a CEO essay as settled policy.

Sep 07, 2026

3 min

Ai innovation

GPT-6 Astra Explained: What OpenAI's Frontier Model Means for Operators

GPT-6 Astra is OpenAI's September 2026 frontier model for hard end-to-end work across coding, computer use, and research. This article unpacks the announcement, ChatGPT and API rollout, pricing and safety constraints, and what engineering leaders should do next.