Writing about the parts of AI and technology delivery that decide whether a thing still works in a year. No vendor pitch, no hype, and nothing that needs a technical background to follow. If you are buying, approving or living with the decision, these are written for you.
The argument about whether AI is reliable enough is the wrong argument. Models and code are two different kinds of machine. The value of this era is not choosing between them, it is engineering the bridge that lets one safely feed the other.
Read the article →The useful question is not whether you trust an AI to make a decision. It is whether the AI is the thing that grants permission. A working trading system shows where the line goes.
Read the article →Four systems at wildly different scales, from filing a receipt to running an autonomous digital company, built on one rule about who is allowed to create a fact. Here is the rule, the evidence, and what it costs.
Read the article →The boss instructs a team. The leader leads one. The introvert works alone and finds it hard going. Then the introvert gets AI and outruns all of them, and that is a match of temperament rather than an accident of tooling.
Read the article →The version everyone posts stops at the panel that flatters them. The two that follow show a leader whose whole team got their own aircraft, and a boss whose team got machines but stayed wired to his. Same technology, opposite outcome.
Read the article →A process that succeeds nine times out of ten sounds like a success. Chain five of those together and you are worse off than a coin toss. The arithmetic behind most stalled AI projects, and the one design decision that fixes it.
Read the article →Every article here comes out of real delivery work rather than research. If something in one of them describes a problem you recognise, the quickest route is a conversation.