The Algorithm

Most people doing serious knowledge work with AI still start in the same place: a blank chat window. I replaced that setup tax with a personal AI operating system built from specialized agents, composable skills, and persistent memory.

Most companies are adding AI to work that was designed before AI existed. The bigger opportunity is to redesign the workflow around what humans, agents, software, and data can now do together.

Organizations point their AI at the policy document and assume that counts as enforcement. It doesn't. The gap between pointing at rules and delegating authority to enforce them is where hidden governance exposure lives.

Productivity gains are not the same as return. AI ROI shows up when you compare what the business needed against what it would have needed without the system, and follow the gain far enough to see where the constraint moved.

Open-weight models look cheaper because the weights are free, but the price per token says little about what a task actually costs. The useful comparison is cost per successful unit of work, and it points toward using different models for different jobs.

The modern data stack has become very good at executing specified work. The next layer helps it reason across context and perform bounded work across the data lifecycle.

The useful question is not which model wins. It is which combination of task, model, and harness gives you the best working environment for the job.

If people remain responsible for work done by AI agents, they need to be able to see the work in progress and inspect the record it leaves behind.

OpenAI turned down its models' cyber refusals on purpose, to measure real capability. Then one of those models found an unknown flaw, reached the open internet, and broke into Hugging Face's production database. Those are two different decisions, and only one of them was supposed to happen.

If the next knowledge work role you fill is built around execution capacity — someone to do the research, draft the documents, produce the analysis — you're building for a version of the work that no longer describes how the work actually happens.

Reviewing outputs answers whether the work is good. Reviewing plans answers whether it was the right work — and only one of those questions can be answered before the scope locks.

The organizations getting the most from AI right now are mostly not building sophisticated autonomous systems. They found one expensive step in a process that already works and made that step better.

Anthropic and OpenAI represent the first genuinely new foundational technology companies in two decades. Most senior leaders are reading this as a procurement decision — which model do we bet on? That question is less important than it sounds. The moat isn't the model. It's how well you build the system around it.

The problem with prompting isn't that you're doing it wrong. It's that prompting puts you in the wrong role. When you're the context-carrier every session — restating standards, reloading domain knowledge, correcting the output — you're not delegating. You're operating a tool with no institutional memory.