Second Opinion, No Stakes
Situation
It was engagement survey week. One of the questions asked me to rate how I felt about my role on a scale of 0 to 10.
I picked a number. Then I got curious.
A number tells you where you are. It doesn't tell you what would move it.
I'd been using a ChatGPT thread as a career coach for months by then, so I took the question there.
The AI Move
I don't have the exact prompt anymore. It was something close to: I need to fill out a survey, the question is how I feel about my role on a 0 to 10 scale, help me imagine what +1 or +2 would look like.
The +1 or +2 was the part I cared about. One or two points I could actually picture.
The Shift
What came back was scary accurate. That's the phrase I used when I posted about it in our AI Slack channel that same week.
It took my professional goals and broke them into tangible, manageable chunks I could act on right away. I don't remember the exact list. It was the size of thing like: set up 30 minutes with a cross-functional partner and have them walk you through their current process, so you understand how their work actually runs.
Nothing on the list was a career overhaul. Every item was something I could start now.
The months of context helped. It already knew my goals and what I'd been working on, so the answer fit my role instead of a generic one. The prompt still works cold, though. That's what I told the channel: drop your number, then ask what +2 would feel like and what's stopping you from moving toward it.
The Pattern
I use a 0 to 10 scale with AI constantly. I ask it to rate its confidence in an answer before I trust it. Then I ask it what we can specifically do to move that confidence level higher.
I act on its suggestions as much as possible because I want to feel confident whenever I'm using an AI response. It's one way I add myself into the AI process. The survey question was the same move, pointed at me instead.
Why +2? It's small enough that it feels manageable, but on a scale of 0 to 10, two points is a fifth of that whole range. That's real movement.
For the engagement survey, the score went into a report. The incremental improvements stayed with me.
The Implication
Pick something you'd rate: your role, a project, a working relationship. Give it a number. Then ask AI what one or two points higher would look like in your day-to-day.
If you don't have a long-running thread full of context, let the AI interview you first. The prompt below does that, and asks for your number at the end.
What I'm Testing Next
I'm working on ways to build the confidence check, and the +2 follow-up, into every AI workflow I run.
I feel like I have AI well-tuned to how I operate, so I'm trusting it more and more. And yes, that means I sometimes miss something it got wrong.
Building that check in would force me to look at where a response needs work before I share it.
The +2 Prompt
I want to reflect on how I feel about [my role / a project / a working relationship]. Start by interviewing me about where things stand right now. Ask one question at a time and wait for my answer before asking the next. Cover what's giving me energy, what's draining it, what I wish I were doing more or less of, and anything that's been bothering me that I haven't said out loud. Ask a follow-up when an answer is vague. Five to eight questions is enough.
Do not give me advice during the interview.
When the interview is done, ask me to rate how I feel on a scale of 0 to 10. Then:
1. Reflect back what you heard in three or four sentences so I can correct anything you got wrong.
2. Tell me what would need to change for my rating to go up by 1 or 2 points, not all the way to a 10.
3. Give me three to five changes. Each one should be small, specific, and within my control, something I could start in the next two weeks without needing anyone's permission. For each change, tell me what it is, why you think it would move my number, and the first step I could take this week.
4. Tell me what is most likely to stop me from following through.
Keep the changes concrete. "Build stronger cross-functional relationships" is too vague. "Set up 30 minutes with [insert name of cross-functional partner] and have them walk you through their current process" is the right size.
If you already have context about me in this conversation, use it, and tell me where it shaped your answer.
Comments