What if you could turn your manager's feedback style, your CEO's decision-making logic, or your favorite expert's frameworks into a reference companion you can consult before making a decision?

That is what I have been experimenting with.

Not using AI to replace people. Using AI to practice better thinking before the real conversation starts.

Example 1: feedback before feedback

Imagine submitting your work to your manager's AI feedback lens before your manager actually reviews it.

It could catch the repetitive issues first: unclear logic, weak structure, missing business context, poor metric definition, or charts that look impressive but do not drive a decision.

Then your actual manager can focus on the higher-value feedback: judgment, nuance, prioritization, and the things AI missed.

Example 2: ask through an expert lens

Another use case is learning from people you admire.

Instead of asking generic AI, "What should I do?", you can ask, "How would this person think about this problem?"

I recently tried this with Avinash Kaushik, whose newsletter I have followed for years since my early analyst days. His writing has shaped how I think about marketing analytics, measurement, KPIs, incrementality, and decision-making.

The problem I kept running into was simple: I remembered a great point from one of his newsletters, but could not easily find the exact issue again when I needed it.

So I built an unofficial AI reference companion using his public articles and my personal TMAI reading notes. The goal is not to replace Avinash or TMAI. It is the opposite: help me ask a marketing measurement question, get an Avinash-style thinking frame, and quickly find the relevant newsletter issue or public article I should go back and read.

  • Not replacing expertise.
  • Making expertise easier to reference, practice, and apply.

Why this matters

I imagine many CMOs, VPs, analysts, and marketing professionals have a similar problem: we remember sharp ideas from trusted experts, but cannot always retrieve them at the moment of decision.

This is where AI gets interesting to me. It can help preserve the connection between a problem, a thinking framework, and the original source you should revisit.

Important boundaries

  • Use public or permissioned material.
  • Do not impersonate the person.
  • Do not expose private content.
  • If the material includes company-private information, keep the tool internal.
  • If you distill a book, newsletter, or paid content with someone's intellectual property, use it only for personal learning.
  • Cite and point back to original sources.
  • Make the tool a thinking companion, not a replacement.

The workflow

  1. 01 / Gather contextUse public writing, permissioned notes, rubrics, talks, books you have legal access to, or your own work history.
  2. 02 / Extract patternsIdentify repeated frameworks, questions, decision rules, feedback style, and source references.
  3. 03 / Build the lensTurn the context into a small agent, prompt, or skill that answers through that specific perspective.
  4. 04 / Keep the loop honestUse it to improve drafts and decisions, then return to the original person or source for deeper judgment.

I am still experimenting, but I think this could become a practical AI workflow for managers, consultants, analysts, and anyone trying to make better decisions.

Distill the thinking. Respect the person. Use AI to think better, not to make humans smaller.

Try the SME Thinking Room Avinash prototype Open-source examples Subscribe to TMAI