The real skill behind good AI prompts is older than computers
Context is the AI gospel right now. But the skill behind it can't be easily taught.
The current AI gospel is simple: context is everything. Give the model enough background, and it produces something useful. Skip the context, and you get something generic.
That’s true. But it skips a step.
Context comes from attention
Giving good context isn’t a prompting skill. It’s a noticing skill.
To give an AI useful context, you have to know what matters about the task in front of you. You have to know what to include, what to leave out, and when to do each. None of that lives in a prompt template. It lives in how closely you’re paying attention — to the work, to yourself, and to what the AI just gave you.
The skill people mistake for something else
Most of the time, our attention is captured and managed for us without any conscious effort. Notifications, ads, half-finished tabs, the next thing on the list. We don’t direct our attention. It directs us.
When people struggle with AI, they usually think the missing piece is technical. A better prompt. A newer model. A workshop. It’s almost never that. The missing piece is the same thing that’s been missing in most knowledge work forever: paying attention to what’s actually happening.
It’s nurtured, not taught
This is the part people don’t want to hear. Attention isn’t a separate skill you pick up in a class. There’s no syllabus. No certification. You develop it on your own, over time, by being curious about what’s in front of you.
You can lead a horse to water. You can’t make it drink. And when it comes to attention, you can’t even lead someone to it. They have to lead themselves.
What this looks like for lawyers
The lawyers I see getting better with AI fastest aren’t the ones who collect the most prompts. They’re the ones who pay attention as they go. They notice what worked last time. They notice when the output drifted. They notice the small detail they forgot to give the model, and they add it next round.
Over a few months, that habit compounds. The scattered, fire-and-forget user is still getting mediocre output a year later. The attentive one is doing things their colleagues don’t think are possible.
Bottom line
The AI gospel is right that context matters. It just stops short of the real point. Context is downstream of attention, and attention is something you have to choose for yourself.
;-)
Ernie
P.S. In the Inner Circle, we slow down together — share what’s working, share what isn’t, and pay attention to what the tools are actually doing.
→
https://ernietheattorney.net/