I finally did something I'd been putting off. I opened ChatGPT and asked it who should teach non-technical professionals in Montreal how to use AI.

It came back with a map, star ratings, four training companies and two universities. All solid options.

Then I asked what it knew about Cinz.AI. It found my LinkedIn, my MBA, and the story of how this newsletter started. It was kind and thorough, and then it told me it wouldn't recommend me yet, "not because I found something negative, but because the public evidence isn't yet dense enough."

Whether you run a business, work in sales, or are job hunting, the reality is that many people will be using AI to learn more about you.

In a US survey this spring, 45% of consumers said they'd used AI tools to find a local business. A year earlier it was 6%.

So where does the AI get its answer? Picture a hotel concierge.

A guest walks up and asks where to eat tonight. The concierge doesn't hand over the phone book (do they still print phone books? 😆). They think of the places they've heard about for years, flip through this week's guide, and give the guest three names with a reason for each.

That concierge is your AI. Here are the four things it does to build an answer about you.

1. It starts with what it remembers. Before it was released, it read a huge chunk of the internet. That memory decides which names come to mind first, and it only updates when a new version comes out.

2. Then it searches. For questions about right now, ChatGPT and most other tools look things up live: your website, your LinkedIn, reviews, directories, articles that mention you.

3. It reads, then picks. It doesn't give you ten links or multiple pages to choose from like Google does. It hands back a short list, with reasons.

4. It shows its sources. When it searches, it lists exactly what it read. That list is the most useful part of the answer.

That's the big difference with Google. Google hands you the list of options and lets you choose. The AI chooses for you, and there's no page two.

And a concierge never says "I don't know that place." If it can't find you, it just names someone else.

So can you change what it says? Not directly, because there's no "claim this profile" button. But you can change what it reads.

When ChatGPT described Cinz.AI, it said we were eight issues in. This is issue 18. It was reading an old LinkedIn post, and that's fixable. Update the pages you own, get mentioned on pages you don't, and expect the search side to move faster than the memory side, which waits for the next version. If it's saying something wrong about you personally, OpenAI has a removal form. It's all in the guide below.

I’m sure you've googled yourself at some point in your life or maybe you even have an alert set up. But when was the last time you looked yourself up on AI? The quick check takes about ten minutes.

1. Open a clean chat. Your AI knows you, and it will flatter you. In ChatGPT, open a new chat, select Temporary, then choose Unpersonalized before you type anything. In Claude, click the ghost icon at the top of a new chat.

2. Ask like a stranger. Leave your name out. Type what someone would ask if they needed you and had never heard of you, like "I'm in Montreal. Who should I hire to [what you do]?" Try two or three versions.

3. Ask for the diagnosis. Paste this in:

4. Try one more tool. Gemini, Claude or Perplexity. They won't agree, and that's useful.

Then look at where you landed. You're named, you're invisible, you're described wrong, etc. Each one comes with a next step, and the diagnosis prompt writes it for you.

🙊 Ads in ChatGPT can now talk back. OpenAI is testing Sponsored Agents, a new kind of ad inside ChatGPT. Click one and you can chat with an AI version of that business, ask it questions, then head to their site. OpenAI says these chats are clearly labeled and kept separate from ChatGPT's own answers. For now it's a small test with selected US advertisers..→ Read it here

🔎 In case you missed it: LinkedIn now finds people by description. LinkedIn search used to work like a filing cabinet: job title, company, city. Now you can describe who you're looking for, like "people who co-founded a productivity startup and are based in New York," and LinkedIn's AI finds the profiles that match, with a short note on why each one came up. It reached all US members this spring, and LinkedIn hasn't said when it's coming to Canada. When it does, your About section stops being a formality. It's what the AI reads.→ Read it here

🛍️ Amazon blocked Meta's new AI from shopping on its site. Meta's Muse is an AI agent that runs errands for you, shopping included. It launched on September 8 and hit the top of the free iPhone apps within a week. Twelve days in, Amazon blocked it, saying unauthorized AI agents break its terms. Meta says Muse can't see people's passwords or payment details. Amazon has blocked shopping agents from Perplexity, Google and OpenAI too. It's this issue's question, one level up: when an AI does the choosing for you, who decides what it's allowed to see?→ Read it here

The line from ChatGPT that stuck with me wasn't the part where it passed on me. It was this: Cinz.AI needs to "make its existing credibility legible to a stranger."

I think that's true for most of us. You probably don't need to be better at what you do. You need to be easier to find, and that part is fixable.

Now the ask. Run one stranger question tonight and hit reply with one word: named, invisible, wrong, or beaten. Mine was invisible (for now 😎)..

Cinz.AI

Your unfair advantage, one week at a time.

Got a question, a topic request, or just want to say hi? Just reply to this email. I read everyone.