AI & Implementation
-
The Lowe’s Tool Shelf Problem
Three new coding agents will ship between morning coffee and afternoon standup, and the reflex to try each one is the trap. Contractors do not own every tool. They own the tools that match the jobs they take.
AI & Implementation · 4 min -
“Tokenmaxxing” Is Lines-of-Code Thinking for the Agentic Era
Engineers now compete on leaderboards for tokens consumed. I have seen this before, with lines of code. Measure the most visible input and you get more of the input and less of the outcome. The metric changes. The mistake does not.
-
Your AI Notetaker Is the Biggest Security Decision You’re Not Making
An AI meeting recorder captures your standups, incident reviews, and board prep, then stores them on infrastructure you do not control. Most teams never decided to allow that. It is one of the biggest security decisions you make this year.
-
Minutes Added To Workforce
Most companies measure AI in cost: jobs not hired, hours saved. That misses what is happening. AI adds minutes to the workforce you already have. The question is whether you track that new capacity and use it on purpose.
-
Block Just Cut 3,500 Jobs. You’re Reading It Wrong.
The headlines say troubled company. I read it differently. The cuts correct a hiring logic from the zero-rate years that was never real, and they go deeper than a return to 2019. Every company built in that window faces the same math.
-
Are You Managing Your Manager? (The Agentic Update)
In 2015 I gave engineers a formula for their careers: dreams, alignment, goals, feedback. It assumed execution was your currency. Agents changed that. Now you have to show that you know what to build and why it matters.
-
Can AI Solve the Four Burner Problem?
The Four Burner Theory says success means turning off family, friends, health, or work. I have lived it. AI may be the first technology that changes the math: more gas for every burner, not a better way to ration it.
-
The Next Platform Giant Is Hiding in Plain Sight
Every era of software produces a platform that becomes the system of record. The talk today is about which coding agent writes the best code. Agents are the editors of this era. The platform is what happens around them.
-
Awe in an AI World
The talk about AI swings between hype and dread. I am stuck on something simpler: people are about to get a lot of their time back. The awe is in watching them rediscover what they are good at.
-
Why Neuro-Symbolic AI Matters Today
Neural networks learn patterns. They do not reason, and they produce confident nonsense when pressed. Neuro-symbolic AI pairs that learning with explicit logic and rules. Sometimes the right answer is a smaller neural component that works with knowledge you can inspect.