💯 What changes when ChatGPT has ads

Window Into the Future 🔮

ChatGPT is getting ads. Here's what I'm thinking about.

You've probably seen the headlines. OpenAI announced a new $8/month tier called ChatGPT Go, and they're starting to test ads in the US.

If this newsletter is your main source of AI stories, genuinely flattered, happy to be your filter. If you already saw it, stick around. The more I dug into this, the more I realized the interesting part isn't the ads themselves.

The basics

ChatGPT Go is $8/month, now available in 171 countries. More messages, longer memory, more image generation than free. Ads are coming to both free and Go tiers. Plus ($20), Team, and Enterprise stay ad-free.

Ads show up at the bottom of responses, clearly labeled, separate from the answer. OpenAI says they won't influence what ChatGPT tells you. Someone on Reddit has a different vision:

The part I keep thinking about

This isn’t really a story about ads but about distribution.

ChatGPT has 800+ million users. 68% of them are on the free tier. That's a massive audience that's about to see sponsored content woven into their AI experience not as banners they can ignore, but as recommendations at the bottom of answers they're actively trusting.

For comparison, contextual ads in AI search (like Perplexity) are getting click-through rates of 3.2%. The display ad industry average is 0.5%. People click on AI recommendations because they feel like advice.

What this means for work

1. Showing up is getting redefined.

If you're in marketing, sales, or building any kind of product, the game is shifting.

Google trained us to think about keywords (eg: rank for X term). AI is evaluating conversational intent. Someone doesn't search "project management software", they ask "I need something to help me manage a remote team of 5 designers who keep missing deadlines."

Microsoft is already calling this shift Answer Engine Optimization. It's less about SEO and more about having documentation that clearly explains what you do, content that answers questions for a machine trying to help someone.

2. AI is becoming the "middle layer" and that changes the transaction.

What’s crazy is OpenAI isn't just adding ads. They're adding commerce. "Instant Checkout" lets you buy things, book Ubers, hire service providers all without leaving ChatGPT. The AI handles the comparison AND the transaction. Think about what that means.

The brand's website gets cut out entirely. So this is the question to sit with: do you partner with these platforms (and risk losing the direct customer relationship) or stay out (and risk being invisible)?

This isn't new, it's the same dilemma Amazon sellers have faced for years. But it's expanding into a new arena.

3. Judgment becomes the skill.

If AI responses start including sponsored recommendations even clearly labeled ones, the skill isn't "avoid AI." It's knowing when to trust what it tells you.

A study found that when models like GPT-4o integrated ads, users often didn't recognize them as ads. They rated the products positively because the recommendation felt like advice. That's not because people are dumb but because we process these conversations as dialogue, not marketing.

So if ChatGPT recommends a specific tool or vendor, maybe verify with a second source. Or ask it "Are there alternatives you didn't mention?" Treat it like advice from a smart friend who might have a sponsorship deal you don't know about.

Bottom line

I guess we're one step closer to this Black Mirror episode.

The Vibe Coding Games are here 🏅

Speaking of building with AI, here’s something I wish existed when I started learning to build with AI.

The Vibe Coding Games kick off February 4th and run through February 25th. It's basically the Olympics of building with AI except you don't need to be a developer to compete. The premise is simple: with AI, anyone can build apps now. You don't need to know how to code. You just need an idea and the willingness to figure it out as you go.

Are you game? Join thevibecodinggames.com 

This week in keeping up with AI 🤖

If you want to dig deeper, here are the sources I used to write this newsletter:

• OpenAI's full announcement on ChatGPT Go and ads.

• AI Agents & Ecommerce, the next AI business tussle.

• Who Will See Ads on ChatGPT and Who Won’t? (Updated 2026)

• Ads that Talk Back: Implications and Perceptions of Injecting Personalized Advertising into LLM Chatbots

🔴 LIVE THIS WEEK: If you're coming from a non-technical background and interested in learning about Claude Code, join Harold and James this Tuesday, Jan 27 for a free session + a live demo of Claude Code 101.

Before you go ✌️

One visual worth seeing

This is where OpenAI says they're heading.

• Level 3 is "Agents - systems that can take actions." That's the Instant Checkout stuff.

• Level 5 is "AI that can do the work of an organization." The ads aren’t just about monetization. It's funding the path to Level 5.

Which level do you think we're at? And which one scares you?

See you next week! 👋