Blog
Thoughts on governed AI orchestration, human-AI workflows, and building systems that scale.
-
RPA Is Running on Borrowed Time
UiPath went public at $56 per share, peaked near $90, and has since collapsed to the $12-20 range. When the market leader is running from its own category name, the category has a problem.
-
Dan Schulman Is Half Right. The Other Half Is What SMBs Need to Hear.
Verizon's CEO predicts 20-30% unemployment from AI. Nvidia's CEO says the opposite. Both are talking about someone else's problem. Here's what actually matters if you're running a small business.
-
Bolt-On Governance Is Expensive Observation
The AI governance market is booming. Credo AI has raised over $25 million, Arthur AI over $60 million. The market exists for a real reason — but the approach is backwards. Governance that monitors AI after the fact is the same architecture that failed in data quality for two decades....
-
The Frankenstein Stack: How Enterprises Are Assembling AI Wrong
Five categories. Five vendors. Five governance stories. Five audit trails. Nobody planned this stack. It assembled itself — one procurement decision at a time, each solving the problem in front of it, none considering how it would all work together.
-
AI Agents, Agentic AI, and AI Orchestration: A Definitive Guide
Everyone is talking about AI agents. Nobody agrees on what the terms mean. 'AI agent,' 'agentic AI,' and 'AI orchestration' are used interchangeably — but they're fundamentally different things. The confusion leads enterprises to buy the wrong solutions for the wrong problems.
-
So We Built the AI Dial
Over five articles, I argued that organizations need the AI Dial for AI autonomy — not a switch. That governance should be built in, not bolted on. That the data quality industry already proved bolt-on doesn't work. This is what we built.
-
I've Seen This Movie Before: What Data Quality Taught Me About AI Governance
Poor data quality cost the US economy $3 trillion per year. The entire data quality industry existed because enterprises built systems first and bolted on quality later. Now the same pattern is repeating with AI governance — and the costs could be worse.
-
When You Work With AI Well, Governance Takes Care of Itself
Most enterprises treat AI governance as a separate project — a layer bolted on top of their AI deployment. But when you delegate to AI well, you've already built the approval workflows, escalation paths, and audit trails that governance requires. You just didn't call it governance.
-
The AI Dial, Not the Switch: Why AI Autonomy Shouldn't Be Binary
Most AI platforms give you two options: off or on. But no organization actually wants binary automation. You want AI to auto-process routine invoices but flag unusual ones. You want AI to respond to simple tickets but escalate complaints. That's not a switch. That's the AI Dial.
-
The Delegation Model: What Working with AI Looks Like in Practice
You wouldn't hand a new hire signing authority on contracts on their first day. Not because they're incompetent — because trust is earned through demonstrated performance. AI is the same. The management science on effective delegation maps directly to how organizations should deploy AI.
-
Working with AI, Not Just Using It
The companies getting real results from AI aren't the ones using it the most. They're the ones who understand where to delegate and where to stay involved. BCG/Harvard found a 40% quality improvement — or 23% worse — depending on which side of that line you land.
-
Your Team of 5, Working Like 50: How SMBs Are Using AI to Multiply, Not Replace
You don't need to hire 10 more people. AI can multiply the capacity of the team you already have — without complexity you can't manage. Here's how SMBs are recovering 40+ hours per week.
-
The 94% Gap: Why AI Isn't Deployed and What Enterprises Are Missing
AI could handle 94% of computer tasks. Only 33% are automated. The 61-point gap isn't a technology problem — it's an adoption infrastructure problem. Here's what enterprises need to close it.
-
Yes, Jobs Are Changing. Here's What's Actually Happening.
Anthropic's landmark labor market study found AI could handle 94% of computer tasks — but only 33% are actually automated. The gap tells us everything about what's really happening with AI and jobs.
-
Every Enterprise Wants AI Agents. Only 5% Can Actually Deploy Them.
79% of enterprises have adopted AI agents. Only 5% made it to production. The bottleneck isn't capability — it's governance. Here's the data, the regulatory reality, and what the 5% do differently.
-
The OpenClaw Moment Has Evolved: AI That Automates Anything Is Here
OpenClaw proved AI can automate anything. The question isn't capability anymore — it's how your enterprise harnesses that power responsibly.
-
Govern Any AI Agent in 5 Minutes: A Technical Guide
Your team is using AI agents — OpenClaw, Claude Code, LangChain, custom tools. They're automating incredible things. This guide makes them enterprise-ready in 5 minutes.
-
OpenAI Just Validated the Autonomous Agent Category — Here's What It Means
When OpenAI hires the creator of the fastest-growing GitHub project ever and backs its transition to a foundation, they're not just making a hire. They're placing a bet on the future — and exposing a governance gap nobody's closing.
-
Why Enterprises Won't Buy Your AI (Yet)
Your AI is impressive. Your governance story is missing. Here's what enterprise buyers actually need to see — backed by procurement data and win/loss analysis.
-
AI Magic Has a Shelf Life
What feels like magic today becomes technical debt tomorrow. Governance isn't friction — it's what separates toys from tools.
-
The 1% Problem: Why 99% Accurate AI Isn't Good Enough
At 1,000 decisions per day, 99% accuracy means 10 disasters. Governance is how you catch them before they ship.
-
Your AI Demo Is Lying to You
The gap between 'it works in the demo' and 'it works in production' is called governance — and the data says it's killing 46% of AI projects before they ever ship.
-
180,000 Developers Gave an AI Agent Root Access. Nobody's Governing It.
OpenClaw hit 180K GitHub stars. Developers love it. Security teams can't sleep. Here's the governance layer that lets you have both.
-
Introducing AICtrlNet: AI Orchestration Where Humans Are First-Class Citizens
We're launching AICtrlNet — an open core AI orchestration platform that treats humans and AI as equal workflow participants. The Community Edition is MIT licensed and available today.
-
Why AI Governance Can't Be an Afterthought
Companies are racing to deploy AI agents. The smart ones are building governance in from day one.
-
The Missing Piece in AI Orchestration: Humans
Every AI orchestration platform focuses on connecting agents. Almost none focus on connecting humans.
-
The Protocol Wars: MCP, A2A, and Why Humans Are Still the Missing Piece
As AI protocols battle for dominance, we're missing the most important integration point of all—humans.
Subscribe via RSS