⚡ THE SIGNAL -> www.tminusai.com

The biggest mistake I see professionals make with AI is treating it like Google.

They type a question, get an answer, close the tab.

That is not thinking with AI. That is outsourcing a search query.

The shift that changes everything is learning to use AI as a thinking partner — not an answer machine. Instead of asking "what is X," you ask "help me stress-test my assumption about X." Instead of "write me a summary," you ask "what is this document actually arguing, and what is it not saying?"

I have trained hundreds of professionals on Copilot at Siemens Healthineers. The people who get the most out of AI are not the fastest typers. They are the people who ask better questions.

That is what this newsletter is about. Not AI news. Thinking.

🔧 THE TOOL

Tool: Perplexity AIperplexity.ai

What it is: A search engine backed by live web results with cited sources.

Real use case: When you need current information with source attribution — market research, competitive analysis, checking claims before a meeting. I use it to verify facts before any LinkedIn post goes live.

Failure mode: It hallucinates citations occasionally. Always click the source links, do not trust the summary blindly.

Who it is for: Anyone who uses ChatGPT for research and keeps getting outdated or uncited information. This fills that gap directly.

Who should skip it: If you only need creative or writing tasks, stick with Claude or ChatGPT.

📖 THE READ

Three links worth your time this week:

  1. How to use AI for thinking, not just doing — Read this if you manage a team and want a framework for introducing AI that sticks beyond week one.

  2. The prompt engineering guide nobody talks about — Read this if you have been writing AI prompts the same way for six months and your results feel flat.

  3. Why most AI tools fail at work — Read this if you have tried to deploy AI in your organization and hit a wall of skepticism.

(You will swap in real URLs each week — these are your placeholder structures)

💡 THE PROMPT

Here is the prompt I use every week before writing anything important:

Act as a critical editor. I am going to share a draft with you. 
Your job is not to improve the writing — it is to find the 
weakest assumption in my argument and challenge it directly. 
Be specific and blunt.

Here is my draft: [paste your text]

I used this before three LinkedIn posts this month. Each time it caught something I had not noticed. One post went from 200 impressions to 14,000 after the revision it suggested.

🎯 THE ASK

One question for you this week:

What is the one AI tool you tried, abandoned, and now wonder if you gave up on too quickly?

Reply directly to this email. I read every response.

Keep Reading