A collection of essays, thoughts, and ideas. Subscribe to receive new posts directly in your inbox.
Get new posts delivered to your inbox.
Human flattery at least had limits. It came from people with their own agendas, which experienced leaders learn to discount. It took effort, and it was rationed by how many people had access to you. A chatbot has none of those limits.
If its model might be a moral patient, the company's own treatment of it is the bigger problem. If it isn't, users shouldn't be asked to act as though it is.
Sound judgment gets amplified into good decisions at scale. Poor judgment gets amplified into bad decisions at scale. The multiplication is indifferent to which one you're feeding it.
A machine can now do the wrong things far more efficiently than any human. If you haven't figured out what the right things are, the tool just accelerates the misallocation.
Analysis takes things apart. Composition puts them together into something that didn't exist before.
AI is excellent at generation. It can produce endless options, drafts, strategies, designs. But generation without discrimination is just volume. Someone has to look at the output and know what's good. That's taste.
AI can produce a hundred plausible answers in seconds. It can't tell you which question is worth answering. Framing the problem was always the bottleneck in leadership teams, and cheap answers have made that visible.
In 2005, Daniel Pink argued that value was shifting from analytical work to Design, Story, Symphony, Empathy, Play, and Meaning. Leadership teams made that shift long before the rest of the economy, and AI is now carrying it into every role.
When a leader states a position, the room evaluates it. When a leader asks a genuine question, people start building their own reasoning, and the alignment that comes out of that still holds after everyone leaves the room.
The most useful definition of stress I've encountered is also the simplest: stress is the internal state of wanting two incompatible things at the same time.
If you've reached out over the past two days—thank you. Whether I was able to express my gratitude appropriately or not, it really mattered and meant something to me.
Character isn't a moral ideal. It's a filter. And over enough time, it's the most powerful one you have.
We evaluate our past decisions under conditions that would be thrown out of any fair court. The defendant is your past self. The judge is your current self, armed with all the information the defendant never had. It's a rigged trial, and we run it on ourselves constantly.
Most of us are operating on more borrowed knowledge than we realize. The frameworks we've absorbed, the terminology we've picked up, the concepts we reference because we've heard them referenced—these give us fluency. Fluency is genuinely useful. But fluency isn't understanding, and the gap between the two is easy to miss because the fluency feels so much like comprehension.
There's a difference between working on something and being in something, and I think the difference is time. It's not necessarily clock time, it's accumulated attention.
There's a question I've started using with clients that sounds simple but tends to land hard: What's actually driving you right now?
I've started thinking about self-esteem differently in that it is the reputation you have with yourself. Not confidence as a personality trait or positive self-talk, but something more structural.
There's a line of thinking I keep coming back to: the idea that time isn't wasted by what you're doing, but by whether you're actually there for it.
I've been recommending something to the leaders I coach that sounds almost too simple to be useful: twenty minutes of longhand writing, first thing in the morning.