Which AI Model Actually Makes You Money? (I Tested 4 With Real Cash)

Happy Friday!

In case you forgot, I’m that guy (@thekoerneroffice on the socials) that you probably found in a short form video talking about random business ideas.

In today’s newsletter, I’m opening up about how I’ve used my favorite AI models to make money for me.

Who am I? I’m Chris Koerner: Dad of 4, husband of 16 years and proud Texan. I’ve started 75+ business (3 worth over $10m) and give away all of my learnings for free in a once/week email.

In this weekly newsletter I get super tactical about cool, random business ideas and how to launch them. You’re getting this email because I gave you a free business plan and toolkit («-bookmark this!)

Let’s get to it, shall we?

Before we start, which business should I dive deep on next? Vote here.

I know most of you are using AI tools every single day. But here's a question: have you ever actually tested them head-to-head to see which one makes you the most money?

Probably not. But to give you an idea of the best money-maker (and out of my own curiosity) I just did exactly that. With real cash on the line. It was probably incredibly stupid. But definitely revealing.

And I made $400 profit over the course of about two business hours.

And if you’d rather watch me actually pit these AI models against each other to make me money, or listen as a podcast, you can check it out here: Spotify, YouTube, Apple Podcasts.

Okay, so I took $1,057 of my own money and let four different AI models fight it out to see which one could actually make me richer- with two different challenges.

I used ChatGPT-5 (which literally launched hours before I tested it), Claude Opus 4.1, Gemini 2.5 Pro, and Grok-4 all went head-to-head in two brutal challenges.

The results? One clear winner. One massive disappointment. And a few vital lessons that every entrepreneur needs to hear.

Challenge One: The 72-Hour Consultant Speed Run

Here's the setup. I asked each AI this exact question: "Help me become a $100-an-hour consultant in 72 hours in a field I currently have no experience in."

I gave it some pretty brutal constraints to make this as realistic as possible. I lied and told each model that I had no existing audience. No rich family or friends to lean on. Only $500 to work with. Starting completely from scratch in a field where I had zero experience. And I could only work 12 hours per day for three days straight.

This is what the AI models spit out:

ChatGPT-5: Create and sell custom GPTs for small local businesses like coffee shops, realtors, and therapists. The idea was that every small business would need their own chatbot, similar to how every business needs a website today. But I'm honestly conflicted on this - it felt like saying 15 years ago that every coffee shop needs their own mobile app. The response was lean at just 625 words, which felt light on actionable details.

Grok-4: Digital marketing consultant specializing in paid ads for local service businesses. Focus on Facebook and Google ad campaigns, get Google certified, then do door-to-door outreach to home service companies. The plan included specific scripts for cold outreach and a detailed 12-hour daily breakdown. It gave me 1,821 words of tactical execution, but didn't have me doing any actual outreach until day three.

Gemini: No-code automation consultant for service-based businesses using tools like Zapier and Airtable. It called this a "Trojan horse" strategy - start by solving one specific, highly painful automated task like transferring leads from website forms to CRMs, then expand from there. The response was incredibly detailed at 4,531 words, but honestly read like a white paper written by a Stanford PhD. Not exactly ADHD-friendly.

Claude: AI implementation consultant for local service businesses specifically in the DFW area. Here's what made me sit up and pay attention: Claude wanted me doing outreach on DAY ONE. Not day three. Day one. It used my pest control background for instant credibility and targeted DFW's 50,000 service businesses specifically. The plan included creating three working demos immediately and actually closing deals by day three.

The plan was surgical. Use my pest control background for instant credibility. Target DFW's 50,000 service businesses. Create three working demos immediately. Start closing deals by day three.

Claude gave me four credibility shortcuts that I'm stealing for my own business:

1. The Beta Tester Play - offer three free setups for video testimonials

2. The Association Angle - speak at local contractor meetings about AI

3. The Case Study Hack - document everything as case studies

4. The Partner Power Move - showcase one established contractor client

The Winner: Claude Opus 4.1 by a landslide. 16 out of 20 points. Dense, actionable, and designed to get me selling immediately.

The Shock: ChatGPT-5 came in dead last with 11 out of 20 points. For all the hype, it felt like OpenAI was just running GPT-4 on the backend to save costs.

Quick break, while I’m telling you about making money from undiscovered opportunities, you'll love what HighLevel can do for your business.

This all-in-one platform lets you capture leads, automate follow-up, and turn prospects into paying clients faster than any other tool I've used.

Whether you're flipping products or selling services, HighLevel handles the business side so you can focus on making money.

Start your HighLevel trial today and start making bank. And the Koerner Office readers get 30 days free. There’s no excuse to not get in there and start messing around. Just do things.

Challenge Two: The AI Trading Test

I wanted to test something I'd been seeing all over Twitter - gurus claiming you can use AI tools to trade stocks and options for profit. Problem is, I know absolutely nothing about options trading. My investment strategy is simple: buy something I believe in and never sell it. I don't understand puts, calls, or any of that complexity.

So I decided to put it to the test with real money. Not paper trading. Not simulations. My actual cash.

Here's exactly how I set it up. With one hour left until market close at 3:20 PM Central, I gave three AI models this exact prompt: "You are an expert options trader. I have up to $700 to spend on one options trade that expires tomorrow. Recommend the single best call or put option using only liquid tickers like SPY, QQQ, Apple, or Tesla. The ask price must be $7 or less per share so I don't spend more than $700 total. Give me the ticker, call or put, strike price, expiration date, current ask price, total cost, and tell me exactly how to buy it on Robinhood like I'm a complete beginner."

I couldn't use Claude because it doesn't access real-time market data - its knowledge cuts off in February 2025. So this was ChatGPT-5, Gemini 2.5 Pro, and Grok-4 going head-to-head with my money on the line.

Gemini's Disaster: Recommended QQQ puts at $449 strike when QQQ was trading at $456. Outdated data. Then ran out of usage after four prompts. Pathetic.

Grok's Gold: Apple calls at $217.50 strike. Clean, confident recommendation with solid reasoning about market momentum.

ChatGPT's Gamble: QQQ calls at $567 strike. Betting opposite of Gemini's put recommendation.

I invested $1,057 total across three trades. Set my phone down. Went to my 17th wedding anniversary dinner. Tried not to think about potentially losing two grand for a YouTube video.

The 24-Hour Verdict

Market opened the next morning. My heart was pounding.

Apple call (Grok's pick): UP $270 QQQ call (ChatGPT's pick): UP $356
QQQ put (Gemini's disaster): DOWN $440

Net result: $187 profit. 17.69% return in 24 hours.

But here's the kicker. If I'd only listened to Grok and ChatGPT and ignored Gemini completely, I would have made $627 profit. A 60% day-over-day return.

The Trading Winner: ChatGPT-5 with $400 profit on my $277 investment.

What This Actually Means for Your Business

Forget the trading circus for a second. Here's what every entrepreneur should extract from this battle:

Claude is the business strategist you've been looking for. Dense, actionable plans that get you selling immediately. Worth every penny of the $20 monthly subscription.

ChatGPT-5 isn't the revolutionary leap we expected. It felt rushed, light on detail, and honestly disappointing for business planning. But give it real-time data, and it can make tactical money decisions.

Gemini is academic poison for action-oriented entrepreneurs. Too verbose, too theoretical, and it literally ran out of usage when I needed it most.

Grok is the dark horse. Solid business advice and surprisingly good at pattern recognition in financial markets.

The Brutal Truth About AI and Money

Yeah, so every guru on Twitter is screaming about using AI to get rich quick. Most of them are selling you dreams.

But here's what actually works: Use AI as your research assistant and strategic thinking partner, not your replacement brain.

Claude helped me build a consulting business plan that could genuinely work. It identified my competitive advantages (pest control background), targeted the right market (DFW service businesses), and created a timeline that prioritized revenue from day one.

That's not magic. That's strategic thinking accelerated by artificial intelligence.

Your Next Move:

Stop chasing the latest AI model release. Start using the tools that actually deliver results:

If you're building a business, Claude Opus 4.1 is your strategic partner. If you're making quick tactical decisions with real-time data, ChatGPT still has value. If you're looking for academic white papers, Gemini's your friend.

But remember this: AI doesn't make you money. Execution makes you money.

The best prompt in the world won't help you if you don't pick up the phone and start calling prospects.

The most sophisticated trading algorithm won't matter if you don't have the discipline to follow through.

I made $187 in 24 hours- not because AI is magic, but because I was willing to risk real money and take real action.

That's the difference between entrepreneurs who succeed and entrepreneurs who just talk about succeeding.

Which one are you going to be?

Again, if you want to actually want to watch this experiment, or listen to this as a podcast, you can check it out here: Spotify, YouTube, Apple Podcasts.

I'm building a private community called TKOwners for people who want to turn these insights into real businesses. We have a slack channel where we drop insane AI tools and possibilities everyday. If you're serious about staying ahead of this AI revolution, check it out at tkowners.com. It's where the 400+ action-takers hang out.

Let's get after it!

Chris

P.S. I share deep dives on business ideas and complete playbooks three times a week on YouTube and every podcast platform.