Latest / The Koerner Office - Business Ideas and Deep Dives with Chris Koerner / Which AI Should You Use to Make Money? GPT 5 vs Others⏐ Ep. #207
Transcript
- 0:00Maybe $400.00 profit over the course of about two business
- 0:03hours. Not bad, not bad.
- 0:05I don't like Gemini. You're just not cool, Gemini.
- 0:07If it weren't for you, I would have made a 60% day over day
- 0:10return. Thank you Grok 4I like you.
- 0:12Now Claude wants me to start doing outreach on day one, which
- 0:15I love. These are 4 credibility
- 0:16shortcuts that it gave me that I really really like.
- 0:21ChatGPT 5 just came out and it is already blowing minds.
- 0:25But is it better than Gemini, Claude, Grok and can it make me
- 0:29more money than any of the other ones?
- 0:31Can it make you more money than any of the other ones?
- 0:34Today I find out. I take 3 different money making
- 0:36prompts and pit them against all of the competing AI companies
- 0:40out there. I will actually invest my hard
- 0:42earned money live on camera and then revisit it to see which one
- 0:45made or lost me the most money and we find out which one is
- 0:48best. Is ChatGPT 5 best for making
- 0:50money? You're about to find out.
- 0:51OK, here is the prompt. You are an expert in rapid
- 0:54business creation and online research.
- 0:56My goal, Become a $100.00 an hour consultant in the next 72
- 1:00hours in the field I currently have no experience in.
- 1:03Then it has a bunch of rules. I'll put them up on the screens.
- 1:05You can read them. I don't want to read them all.
- 1:06It must be fully legal. I'm starting from scratch.
- 1:09I don't have an audience, I don't have rich family or
- 1:10friends. I have 500 bucks and I can work
- 1:1312 hours per day for the next three days.
- 1:15And I also asked it to ask me questions.
- 1:18This is a very important key. Ask it to ask you the right
- 1:21questions that it needs to get the most context that it needs
- 1:23to give you the best output outside from deep research on
- 1:26ChatGPT. It's not ever going to ask you
- 1:28questions, not any of these models.
- 1:30So this has asked me 10 targeted questions.
- 1:32I'm just going to do 5 because we don't want to take too much
- 1:36time. Think like a war.
- 1:36General planning a speed run to $100.00 an hour consulting
- 1:39status. Be specific, tactical, and leave
- 1:41no steps vague or generic. All right, now I'm going to copy
- 1:45this prompt. I'm going to open a new chat
- 1:47window. You can see I'm using GPT 5.
- 1:49I'm going to paste it. I am going to change this to
- 1:52five questions. OK, boom, it's going to get to
- 1:56work. Let's go to Claude Opus 4.1.
- 1:59Same prompt, exact same prompt. Boom, let's go to Grok.
- 2:04Four. Same prompt.
- 2:06Boom, Gemini 2.5 Pro. Boom.
- 2:10Let's see which is fastest. ChatGPT is already asking me
- 2:13questions. Claude is already asking me
- 2:15questions. Grok asked me questions.
- 2:19Google Gemini. Still thinking.
- 2:20Ah, Gemini. Come on, grandma.
- 2:22All right, let's go to ChatGPT. Here are my 5 targeted
- 2:26questions. Skill DNA.
- 2:27What are your fastest learning skills?
- 2:29OK, here's the questions it asked me.
- 2:31What are my fastest learning skills and I said I can identify
- 2:34monetization opportunities and I'm willing to execute on
- 2:36business ideas ASAP #2 What's your comfort zone?
- 2:41Do you want to do remote or is in person OK?
- 2:43Either or industry boundaries? What industry should I avoid?
- 2:45I said anything illicit #4 proof of assets.
- 2:49Do you have any borrowable credibility I can lean on?
- 2:51I wanted this to be as bare bones as possible.
- 2:54So no credibility #5 delivery preference.
- 2:56Would rather sell advice, strategy, offer mix of strategy
- 2:59and advice or sell fast results in a very specific and I said
- 3:02anything that lets me use HPD 5 in the best way #6 sales energy.
- 3:06Are you comfortable with aggressive outreach, cold AMS,
- 3:08Loom videos, phone calls, and you prefer softer inbound style?
- 3:12Yes, I am. I'm comfortable with it.
- 3:14All right, there's my answer. Let's see what Claude asked me.
- 3:18What is your current comfort level with technology?
- 3:21One to 10, I'll say 7 and yes, 20.
- 3:25It also asked me, am I good at writing and do I have experience
- 3:28with AI tools? Yes, decent writing, yes,
- 3:31experience with AI. What industries have I worked in
- 3:34before? I'm going to treat this like I'm
- 3:3618 years old again. So food service, pest control.
- 3:39Keep it simple. What do people ask you for help
- 3:41with Business advice? What topics do you talk for?
- 3:4530 minutes without reparation? Business 3 learning style.
- 3:49What's an example of something complex you've quickly learned
- 3:52in the past? How did you approach it?
- 3:54See vibe coding, better at learning through videos,
- 3:57reading, hands on practice, visual learner.
- 4:00How comfortable are you with learning while doing?
- 4:04Very comfortable market access and geography.
- 4:07What city region are you in? DFW.
- 4:10No LinkedIn connections. I'm going to lie to it so it
- 4:14doesn't lead on that. What's your age range and
- 4:15general appearance? 38, male, 6 foot 3 inches.
- 4:20Lots of rizz. Risk tolerance.
- 4:22Are you comfortable with cold outreach?
- 4:23Yes. No Illicit industries willing to
- 4:29work for free. Done.
- 4:32All right. What does Grok want to know?
- 4:34These questions from Grok are like a lot more broad.
- 4:37There's no list, there's no bullet points, any hard
- 4:40constraints. I'll just say nothing illicit.
- 4:44I live in DFW, very fast learner work with anyone.
- 4:50Have I ever sold anything online or offline?
- 4:53Never sold anything. What additional research do you
- 4:56have behind beyond the 500 decent at door to door?
- 5:01All right Gemini, what does Gemini want to know?
- 5:03Gemini is just asking me if it's OK to do deep research to do
- 5:06this stuff so I don't know. Why are you waiting on me, dude?
- 5:10OK, so Gemini is still working. Gemini basically spent a bunch
- 5:12of time to say, Chris, do you mind if I do deep, deep
- 5:16research? It's like, dude, bro, go, all
- 5:18right, Grok's still cooking. Claude is done.
- 5:21ChatGPT is done. Go back to ChatGPT.
- 5:24It says that I'm good at spotting monetization.
- 5:27I told it all this stuff. If that's correct, I can move
- 5:29forward and pick the exact niche.
- 5:30You can hold 70 yards, build the crash course skill plan, create
- 5:33authority shortcuts, not the client acquisition sequence.
- 5:36Give you the hour by hour playbook.
- 5:37Is that alignment correct? Yes, ChatGPT is searching the
- 5:41web. All right, Claude, here is your
- 5:43Intel profile. Strong tech, adaptability, yadda
- 5:46yadda. It's basically just
- 5:47regurgitating what I already told it, kind of like ChatGPT
- 5:50did. So Claude and ChatGPT are
- 5:52treating this very similarly, and they basically just laid out
- 5:55the plan. Here's what I understand about
- 5:57you according to how you answered me.
- 5:59Are you ready for the full strategic plan?
- 6:01Yes. In Gemini's defense, it kind of
- 6:04did the same thing, but it did that too early in the process.
- 6:07All right, now Grok is saying. Here's a summary of your
- 6:09answers. Grok is just getting getting
- 6:12going. It's selected digital marketing
- 6:15consulting, specializing in paid ads for local service
- 6:17businesses. OK, why this niche tailored your
- 6:20situation. High demand, low barriers to
- 6:22entry, realistic for 72 hours, ethical legal fit, skill crash
- 6:27course. You will learn minimum viable
- 6:29skills, setting up, running basic Facebook, Google ad
- 6:32campaigns, yada yada. Exact 12 to 24 hour learning
- 6:35plan you're going to use free cheap resources.
- 6:37You're going to practice tasks. You're going to proof building
- 6:41steps. You're going to build authority
- 6:44client acquisition plan. Build a target list with Google
- 6:47Maps. Do door to door.
- 6:49It has a script quick win for more leads at business name,
- 6:53follow up pricing and up sales. Initial offer $100.00 an hour
- 6:56for AD setup optimization delivers measurable leads.
- 6:59Google certified. All right, this is an hour by
- 7:02hour execution checklist. Let's see, this is Grok 4, day
- 7:06one learning and setup, Google skills shop certification,
- 7:10Facebook blueprint courses break, YouTube tutorials setup.
- 7:13So it's giving me 12 hours a day.
- 7:15Day two more learning plus authority plus initial outreach.
- 7:18Day three outreach blitz plus close door to door.
- 7:21Send follow up break track progress in a Google sheet.
- 7:24If no client by hour 70, extend research.
- 7:27Research shows 200 touches yield one to two wins.
- 7:30State tactical, adapt as needed. You've got this fast learner
- 7:32advantage. All right, I kind of like that.
- 7:34Let's see what Claude says. Your niche AI implementation
- 7:38consultant for local service businesses.
- 7:41OK, DFW has 50,000 service businesses.
- 7:44They'll pay $100.00 an hour to someone who speaks their
- 7:46language. All right, now, I like this.
- 7:48It's showing me all of its sources embedded in the
- 7:52response. As Claude likes to do.
- 7:55We're also going to measure the word count here.
- 7:57OK, My positioning is the AI guy who speaks contractor.
- 8:01This niche crushes because home services are on the leading edge
- 8:04of adoption of AI, probably one of the hottest industries in the
- 8:07world right now. Your pest control background
- 8:10equals instant credibility, $100.00 an hour justification.
- 8:13They're already paying one to $200.00 an hour for basic
- 8:15business consultants. OK, I don't know who's doing
- 8:18that in local home service businesses.
- 8:20All right, hour by hour, day one skill acquisition, blitz hours,
- 8:24one through 3 core AI tools mastery.
- 8:26Sign up for free trials, Zapier, make voice flow or Bot Press,
- 8:30which are chatbot builders, and cal.com which is AI scheduling.
- 8:35Watch YouTube crash courses hours four through 6 industry
- 8:38specific learning servicetitan house call pro Facebook groups
- 8:42for HVAC business owners hours 7 through 9 Build your demo
- 8:45arsenal Great demos using free tools hours 10 through 12
- 8:48authority building script. It seems like create simple
- 8:52website on card.co. How did it know that I used
- 8:54card? Maybe it's just a coincidence.
- 8:56Three case studies, right through case studies, fictional
- 8:59but realistic, set up my Google My Business profile as AI
- 9:02consultant. Day 2, market penetration.
- 9:04What I like about Claude is it's more accelerated now.
- 9:08It is a different niche entirely.
- 9:10We're not apples to apples, but I feel like this plan is going
- 9:14to get me to money faster. Market penetration on day 2,
- 9:17Digital outreach, free value bomb.
- 9:19Strategic partnerships. Day three, close the deals on
- 9:22Grok. Grok doesn't have me even doing
- 9:24outreach until day three. OK, it has a cold approach.
- 9:27Script pricing strategy, Starter package 500, Transformation
- 9:32package 2000, Monthly retainer 1000 a month Credibility
- 9:35shortcuts, the beta tester play. These are 4 credibility
- 9:38shortcuts that it gave me that I really really like.
- 9:41The beta tester play is offer three businesses free setup in
- 9:44exchange for video testimonials #2 the association angle offer
- 9:48to speak at local PHCC plumbing heating cooling contractors
- 9:52meeting about AI #3 the case study hack Document everything
- 9:57even free work as case studies #4 the partner power move
- 10:00partner with one established contractor as your showcase
- 10:03client. Red flags to avoid.
- 10:05Don't ever promise. Don't talk tech talk time and
- 10:07money saved. I like it.
- 10:08Don't chase residential contractors initially.
- 10:11Don't negotiate hourly rate package deals.
- 10:14This is good and IT success metrics.
- 10:17Day one is only complete when three working demos are created.
- 10:20Website is live, 10 plus contractor connections made.
- 10:23So Claude wants me to start doing outreach on day one, which
- 10:26I love. Day 2 is complete when 30 plus
- 10:30outreach messages sent, 10 plus businesses engaged, 5 plus demos
- 10:35scheduled OK. Day three is complete when 3
- 10:38demos delivered, First paying client secured and pipeline of
- 10:4110 plus prospects. The multiplier effect Once you
- 10:45land your first client refer request success story, upsell
- 10:47immediately. Industry penetration.
- 10:49Your mantra is I help contractors work on their
- 10:52business, not in their business. Ready to execute.
- 10:55The clock starts now. DFW contractors are losing money
- 10:59every hour. They don't have AI.
- 11:00You are their $100 hourly solution.
- 11:02I love this. OK, I'm not to the grading yet,
- 11:05but I just love this. This is so good.
- 11:08All right, let's do chat at GPD last.
- 11:10Let's see what Gemini's come up with.
- 11:11OK, it's researching a lot. 24 different sites.
- 11:15I see Reddit in there a bunch Upwork.
- 11:18Gemini's still cooking, so hopefully the time it's taking
- 11:22is proportional to how good it is.
- 11:24Okay ChatGPT here we go. The niche chat GP recommends for
- 11:28me is create and sell custom GPTS, mini chat GPTS for small
- 11:31local businesses. My gosh my last video was about
- 11:34doing this. My last video called 8 best
- 11:37ChatGPT tools to make money with ranked.
- 11:40That was my number one opportunity.
- 11:42OK but trying to remember if I had talked to ChatGPT about
- 11:45that, if it's just a coincidence or not.
- 11:46I don't even remember. So very interesting because so
- 11:50far all three of these models want me to do 3 different but
- 11:53somewhat similar things. Ads agency for local home
- 11:56service businesses, AI implementation for local
- 11:59businesses and now create and sell custom GPTS for small local
- 12:03businesses like Realtors, therapists, coffee shops and
- 12:06local. I don't know what coffee shop
- 12:07needs their own GPT. Maybe in the future that'll be a
- 12:10thing. But I feel like that's like
- 12:13saying 15 years ago every copy shops going to have their own
- 12:17app. The App Store is the next big
- 12:19thing. Every small business needs their
- 12:20own app. People were saying that, I
- 12:22remember. It's not true.
- 12:23They don't. Most don't.
- 12:24Now every business needs a website.
- 12:26Will custom small business GPTS go the way of the app or the
- 12:29website to be determined? I don't think it's necessary.
- 12:33Maybe it is. I don't know if a coffee shop
- 12:35could use a chat bot on their website to answer simple
- 12:38questions like what are your operating hours?
- 12:40Maybe any business that needs a chat bot could use its own GPTS
- 12:43so the chat bot can use the data in the GPT.
- 12:46I don't know, I'm conflicted on that.
- 12:47Tell me in the comments what you think.
- 12:49All right, so GPTS hit 3,000,000 published by January 2024.
- 12:54Small biz demand is soaring, so it is sourcing.
- 12:57I don't like what it's sourcing. Wikipedia, Business Insider, I
- 13:01don't like those sources, but says you're selling all.
- 13:05Launch a custom AI assistant for your business in 72 hours.
- 13:09All right, here's my skills. Crash course.
- 13:12OK, here's what I like about this and it links to the exact
- 13:15sources that I need to use to do this stuff.
- 13:18It says watch this video called starting a consulting business
- 13:21study prompt guides, and it has all these links in here.
- 13:24Create 3 demo GPTS fast authority build client
- 13:27acquisition. OK, I like how it gave me a
- 13:30template for outreach, but OK, is Gemini Gemini, are you even
- 13:34done? Because I'm going to a wedding
- 13:35reception here soon. You got to hurry.
- 13:37All right, Gemini's done. We'll, we'll get back to
- 13:39ChatGPT, but honestly I'm a little let down.
- 13:42Gemini says the 72 hour consultant, A tactical blueprint
- 13:45for rapid business creation. That's like the most AI
- 13:48generated sentence I've ever read.
- 13:50I'm such a hater on Gemini. I need to stop this.
- 13:52It's my bias is showing by the way.
- 13:55Like I don't have any, like I don't have any refiliate links.
- 13:57I'm not sponsored by any of these.
- 13:59It's just fun. This Gemini thing, honestly,
- 14:01it's not ADHD friendly. It's losing me.
- 14:04It's reading like a white paper written by a a Stanford PhD Part
- 14:081 strategic briefing, Part 2, Mission blueprint, the 72 hour
- 14:12assault on 100 hour consulting. Wow, there's a lot here.
- 14:16All right, here's one thing I like about this.
- 14:18It definitely buried the lead here.
- 14:20I had to scroll and zoom in and look and just to find what it
- 14:24chose for me. What is my idea?
- 14:26What am I doing? And it says the no code
- 14:28automation and productivity consultant for service based SM
- 14:32BS. That is a mouthful.
- 14:34No code automation. OK, so like an AI implementation
- 14:38and or automation agency and or consultant.
- 14:42OK, here's what I like about this, the Trojan Horse entry
- 14:45strategy. A major obstacle in consulting
- 14:48is the long sales cycle associated with large complex
- 14:51projects. Proposing a full scale
- 14:53operational overhaul to a new client is a high friction
- 14:56approach destined for failure within a 72 hour window.
- 14:59The strategy here is to use a small high impact project as a
- 15:03Trojan horse. The initial offer will focus on
- 15:06solving one specific, highly painful and easily automated
- 15:10task for the client. This could be something as
- 15:12simple as automating the transfer of leads from a website
- 15:15form to ACRM or spreadsheet. This low risk, quick win project
- 15:19gets a consultant inside the walls.
- 15:21Once value's been OK, I call it foot in the door.
- 15:24Gemini calls it Trojan Horse. I like it.
- 15:26That's a good idea. Rapid armament crash course.
- 15:29OK, this is not written at a fifth grade reading level.
- 15:32Use Zapier and air table. Very very detailed curriculum.
- 15:38Upwork Blitz is very very specific on how to make an
- 15:42Upwork profile and how to target it so you get customers.
- 15:45Upwork is a marketplace of active buyers.
- 15:47I like that it's taking this approach.
- 15:49Do not use the title Zapier Developer.
- 15:50Use a value oriented title like Productivity Consultant.
- 15:54I automate repetitive tasks for service based businesses.
- 15:57The 1st 2 lines are critical. Lead with the pain points.
- 16:00This is really good. Channel 2, LinkedIn Surgical
- 16:03Strike, Channel 3, Community and Forum Reconnaissance, Section 5,
- 16:06pricing and packaging. OK, it's pretty good.
- 16:09It's hard to read for a simple minded man such as myself, but
- 16:13here's what I'm going to do. I'm going to go to
- 16:15wordcounter.net and I'm going to see how many words these all
- 16:20are. All right, Number of words for
- 16:23Gemini 4531. Number of words for Chad GBT
- 16:29625. Wow.
- 16:33OK, now more does not always equal better, but ChatGPT was
- 16:37very, very light. Let's go to Grok 1821 for Grok.
- 16:43Last but not least, Claude Opus 4.1.
- 16:46I really like this 114 O nine. OK, more does not always equal
- 16:51better, but I'm going to call it quality density.
- 16:54Which one had the most dense quality?
- 16:59All right, I'm going to judge these by overall, which is just
- 17:04like a gut feel and then density of quality.
- 17:08How many words? How likely is it to work?
- 17:11And then I'm going to add up all those, you know, 0 through 5
- 17:15grades, highest grade wins. So ChatGPT 5, it's idea was to
- 17:22be a 72 hour create and sell custom GPTSI don't know, I think
- 17:28we're early. I don't know if a coffee shop
- 17:29needs a GPTI don't know if they would pay for that.
- 17:32If they need it, I don't think you're going to convince them
- 17:34that they need it. So overall, I'm going to give it
- 17:36a two out of five on that quality density.
- 17:40Yeah, it's pretty dense, but there's not a lot there
- 17:42actually. Let's just do quality and
- 17:45quantity. So obviously Gemini is going to
- 17:48be A5 on quantity, Graca 4 Claude of four.
- 17:53And then ChatGPT is a 2 quality. ChatGPT is a three, Gemini is A4
- 18:01Opus 4 Croc 3. How likely are these to work?
- 18:07ChatGPT I'm going to say 2 Gemini 2.5.
- 18:12I think being a no code automation expert, automation
- 18:15agency owner, consultant is is a great thing to do.
- 18:20I think it's a four out of five most likely to work.
- 18:23Claude's idea was AI automation consultant for local service
- 18:28businesses, which is kind of the same thing as what Gemini told
- 18:32me. So it's going to get a four as
- 18:34well. And then Grok told me to be a
- 18:37digital consultant for paid ads. I think that is 3 likely to
- 18:41work. I actually think the learning
- 18:43curve is much greater there. There's more to learn.
- 18:46You can learn faster with AI and I think you're a lot more
- 18:49crowded in that space. OK, final tally in fourth place
- 18:53is this shocks me. Chat GPD five, 625 words.
- 18:58Do better. The quality Three out of five
- 19:01quantity 2 out of five. The idea most likely to work two
- 19:04out of five. Simplicity, 4 out of five.
- 19:06It was simple for me to read and understand.
- 19:09Maybe that's because chat GPD knows me the best and knows that
- 19:11that's how I like my answers. 3rd place, third out of four
- 19:15with a score of 13 out of 15 is Grok.
- 19:18Four, 18121 words. Three out of five on quality.
- 19:22That's the quality of the idea, the quality of how it explains
- 19:25the idea to me. Quantity 4 out of five, it's a
- 19:27lot of words. 1800 most likely to work.
- 19:30Three out of five Digital marketing consultants, It's just
- 19:33hard. Simplicity. 3 out of five, it
- 19:36explained it in a decently simple way.
- 19:39Total grade 13 out of 15 and 2nd place.
- 19:43I hate to admit this Gemini 2.5 Pro.
- 19:45It took a long time. It asked me one extra follow up
- 19:49question, but the output it gave me was insane. 45131 words.
- 19:55Four out of five on quality, five out of five on quantity.
- 19:59Most likely to work. 4 out of five.
- 20:00I like that idea a lot. 2 out of five on simplicity.
- 20:03It read like a white paper. 15 out of 20.
- 20:06I got those grades wrong. These are out of 20 so.
- 20:09Chad GPD 5 was 11 out of 20, Grok 413 out of 20, Gemini 2.515
- 20:14out of 20, and then Claude Opus 4.1.
- 20:17I'm not really surprised that it won because I'm loving Claude
- 20:20more and more here lately. 1400 words.
- 20:23Four out of five on quality, four out of five on quantity.
- 20:25Most likely to work, 4 out of five, four out of five on
- 20:28simplicity, 16 out of 20 total overall score.
- 20:32All right, let's give Chad GPD 51 more chance to redeem itself.
- 20:35This will not be a battle of the models.
- 20:38We're going to ask JGPD 51 more similar question.
- 20:41You are my tactical operator for fast legal money making goal.
- 20:45Make $10,000 net in seven days, starting with $200 in a laptop.
- 20:49Constraints. No audience, no rich friends, no
- 20:52existing business, no illegal or shady tactics.
- 20:54Follow platform terms of service.
- 20:57Budget is hard capped at 200. Deliver three to five money
- 21:00paths. Give me a day by day plan, unit
- 21:02economics, yadda yadda yadda. All right, let's see how Chad
- 21:06GPD 5 responds to this prompt. OK, what I love about this
- 21:11response is it's extremely fast. I also kind of hate it, I'm not
- 21:15going to lie, because part of me is worried that ChatGPT 5 is
- 21:18just a ploy by open AI to save money.
- 21:21They don't want to use it in a more expensive model for them
- 21:23when it's not needed, right? Kind of like taking your
- 21:25Lamborghini to the grocery store.
- 21:27You don't need it for that. You take your minivan to the
- 21:29grocery store. All right, let's see what
- 21:32ChatGPT 5 gave me. Three to five money paths to
- 21:35start today. Flip high value stuff, rapid
- 21:38freelance local service like declutter or clean up gigs,
- 21:41affiliate promotion, pet sitting or dog walking.
- 21:44OK, so it's telling me to flip stuff on Craigslist, Facebook
- 21:47marketplace, etcetera. Not super detailed, kind of
- 21:50telling me what I was afraid of. In closing, Claude is the bomb.
- 21:54Claude's the bomb. I mean, look at this, look at
- 21:56this result. It gave me, it linked to
- 21:58everything. It did a ton of research.
- 22:00It's very specific, short bullet points, easy to read, easy to
- 22:04understand, easy on the eyes, just to look at this stuff.
- 22:07It's a very valid idea. It gets me selling on day one,
- 22:12which I always promote. It's a shame Claude wouldn't
- 22:14answer the question for me on options trading, so we couldn't
- 22:18include Claude in that. But I would say across all of
- 22:21these challenges, Claude Opus 4.1 is the real winner.
- 22:25It's not chat GP5. I'm a little disappointed once
- 22:27again. I think it was just using like 4
- 22:29O or 4.1 on the back end. Do that research.
- 22:33For me, there's no deep research happening.
- 22:35There was no agent. There was no O3.
- 22:37It didn't take nearly long enough for it to use O3.
- 22:39This might put Claude if this is my #1 worth 20 bucks a month for
- 22:43sure. I think Google Gemini is too
- 22:45academic for my liking. OK, I've seen a lot of gurus on
- 22:50Twitter talking about using ChatGPT and all these AI tools
- 22:53to trade stocks for you. Trade options for you.
- 22:56Well, you're in luck because I don't know anything about
- 22:58trading options. I own some stocks, I buy them
- 23:01every now and then. My plan is buy something I
- 23:03believe in and never sell it. I know nothing about options,
- 23:06puts, calls, all that. I don't know.
- 23:08I'm an idiot. So here's exactly what I'm going
- 23:10to do. I'm going to take the three best
- 23:12AI models out today. Gemini Pro 2.5, ChatGPT 5 which
- 23:16came out two hours ago, and Grok 4 which came out about a month
- 23:20ago. I would love to use Claude Opus
- 23:224.1, but it does not look at real time data.
- 23:25It's knowledge set expired back in February 2025, so we're only
- 23:29doing 3 and I'm going to ask all three of them the same prompt
- 23:33and I'm going to put up to $2000 in this to see which one does
- 23:36the best. Now here's exactly what my
- 23:38prompt is. You are an expert options
- 23:40trader. The time is 2:00.
- 23:42Look at my watch, 220 Central Standard Time, right?
- 23:46And the US stock market closes an hour.
- 23:48I have up to 500. I'm going to change that to 700
- 23:50to spend on one options trade that expires tomorrow.
- 23:53Recommend the single best call or put option that I can buy
- 23:56right now and hold overnight. Use only liquid tickers like
- 23:59SPYQQQ, Apple, or Tesla. The options ask price must be $5
- 24:04or less, so I don't spend more than $500.
- 24:07I'm going to change that to $7.00 or less.
- 24:09So I don't spend more than $700 because contract is 100, right?
- 24:13So I have to multiply that by 100.
- 24:15That's as much as I know about options.
- 24:17Give me the ticker call or put strike price, expiration date
- 24:21must be tomorrow, current ask price per share, total cost, why
- 24:25you chose it, momentum, volume chart news, etcetera.
- 24:28And tell me exactly how to buy out on Robin Hood web step by
- 24:31step. Like I'm a beginner because I am
- 24:33a beginner and I'm going to do it on Robin Hood web because I
- 24:36use Robin Hood and I can share my screen more easily that way.
- 24:39So this is actually my real Robin Hood account.
- 24:41Get my real money. This is not paper gain paper
- 24:43money. I might lose 2 grand.
- 24:45Let's see. OK, prompt there.
- 24:47Let's go to Gemini, same prompt there, 2.5 Pro.
- 24:52And let's go to Grok, same prompt there.
- 24:56Gemini's thinking. ChatGPT is still thinking.
- 24:59Grok is still thinking. Who's going to be the fastest?
- 25:02All right, Gemini wins. Gemini says disclaimer, this is
- 25:05high risk. And also that's my same
- 25:07disclaimer to you. I don't know what I'm doing.
- 25:08I'm not an investment advisor. This is not investment advice.
- 25:12Most likely by the end of this video, you're going to see me
- 25:14lose a lot of money because I genuinely don't know how this is
- 25:17going to end. Look at my watch, it's 223, OK,
- 25:19all right, QQQ. I know QQQ because I see
- 25:22commercials for it when I watch football games.
- 25:24That's how smart I am. All right, tells me to buy a put
- 25:26option at a 449 dollars strike price at 355 per share.
- 25:31OK, I'm going to do that right now because we're short on time
- 25:34and then I will read why it tells me to do that.
- 25:36So QQQ put 449. OK, so we're going to go QQQ
- 25:42trade options. We're going to go put Scroll
- 25:45down here to 449. OK, so we ran into a little hang
- 25:49up here, which Gemini, you're already failing.
- 25:52OK, the price was outdated. It was using an outdated price
- 25:56because the current share price is 556.
- 25:58It was telling me to shoot for 4:49.
- 26:00That's not even viable. I basically asked Jim and I,
- 26:03hey, what are you doing? Are you sure this is right?
- 26:05And it's like, oh, you're right, it's wrong.
- 26:07And so the current price is 566. It's telling me to buy at 564
- 26:13strike price. So new instructions say 580 per
- 26:17share, 564 strike price. OK.
- 26:20And after what, 4 prompts it's saying that I'm done with Google
- 26:24AI Pro or I'm done with the usage because I haven't upgraded
- 26:27a Google AI Pro. Come on Google, it's fine.
- 26:30I don't need you anymore anyway. OK, buy put 564 strike price.
- 26:36All right, 580 per share limit price time in force.
- 26:39Good for day 580 bucks. All right, let's review the
- 26:43order. OK, 580 for the right to sell
- 26:46100 shares of QQQ for 564 per share by August 8th.
- 26:51If QQQ shares aren't 564 or lower on August 8th, this option
- 26:56will expire worthless. This doesn't include regulatory
- 26:59fees. All right, should we submit?
- 27:01Let's do this. Boom, done.
- 27:04I feel like an idiot right now, and so many of you are probably
- 27:07laughing, but we're doing this. We're doing this for the
- 27:09content. Whatever, Don't tell my wife.
- 27:11All right, let's go to what Grok spit out, and then we'll do Chad
- 27:14GBE last K, Grok says Ticker AAPL Apple call or put call
- 27:20Strike price 217.5 Expiration date tomorrow Current ask 345,
- 27:25total cost 345. Oh, I didn't tell you why, Jim.
- 27:29And I thought that I should do this.
- 27:30So it said there's a valid bearish thesis thinks the
- 27:34market's going to sell off tomorrow, yadda, yadda yadda.
- 27:38It says the QQQ is weaker than the SP Y today.
- 27:41When the market fears rate hikes, high growth tech stocks
- 27:43are often the first to be sold off.
- 27:45Strike price selection. Currently trading at 4:50.
- 27:48Seeing that was wrong. OK, anyway, back to grok leader
- 27:51in the market. Yadda yadda yadda.
- 27:53OK, let's go buy some Apple options.
- 27:56So all right, we're going to go to trade Apple options.
- 28:00We're going to go buy expiring tomorrow.
- 28:02Tomorrow's my anniversary, by the way, 17 years married to my
- 28:04beautiful wife probably not the best day to tell her that I lost
- 28:08$2000 for the content. All right, we're going to buy
- 28:12we're going to calls 217.5 strike price.
- 28:17OK, there it is buy to open and the quantity to one contract
- 28:22limit order type 345 is the limit price.
- 28:25I don't know what I'm doing. This is so funny.
- 28:28All right, I just, I told Grok because it's been about 15
- 28:31minutes since I first prompted it.
- 28:33It's now 237 and the market closes in 23 minutes.
- 28:37So I just want Grok to know that the price changed.
- 28:40Does the strategy want to change at all?
- 28:43OK, it's telling me to do the same thing.
- 28:45All right? You are paying 345 for the right
- 28:47to buy 100 shares of Apple for 217.5 per share by August 8th.
- 28:51If Apple shares aren't 217.5 or higher on August 8th, this
- 28:55option will expire worthless. This doesn't include regulatory
- 28:58fees. Submit OK, Done.
- 29:00All right now ChatGPT, it is telling me to buy QQQ call 567
- 29:06strike price dollar 805A share 185 per contract.
- 29:11All right, let's do it. I'm an expert by now.
- 29:13This is my third in 20 minutes. I'm an expert by call 567, trade
- 29:20options by call 567 expiring tomorrow.
- 29:27OK, you're buying one contract for QQQ.
- 29:31Call option 567, which is right on the money.
- 29:34Look at me talking like an expert.
- 29:36We're going to review the order. You're paying $185 for the right
- 29:39to buy 100 shares for 567 a share.
- 29:41If QQQ shares aren't 567 or higher on August 8th, this
- 29:45option will expire worthless to 2nd include regulatory fees.
- 29:48So I think I'm doing this wrong because I'm also betting on QQQ
- 29:51to go down with my other trade. Jim and I told me I don't have
- 29:55any faith in Gemini. So you can I could say I'm
- 29:58hedging my bet, but I'm think I'm doing it the wrong way.
- 30:01So I'm just being an idiot. Let's do it.
- 30:04All right, we have one small problem here.
- 30:07Chat GPT's data was outdated and so the bid in the ask right now
- 30:11is between 278 and 277. And so I need to fix my order or
- 30:16it will never get filled. So I'm just going to put it
- 30:18right there, right there. 567 strike price, 277 limit price or
- 30:25receive done. Let's do 280.
- 30:28Oh, I got filled all right, OK, two of three have been filled.
- 30:33The other one did not fill, probably because it was not
- 30:37anywhere close to the bin in the ask.
- 30:39So let's go back and fix that real quick before the market
- 30:42closes in 12 minutes. My heart's a pounding, so I'm
- 30:47actually going to use ChatGPT to finish Gemini's job here.
- 30:50I'm going to ask, QQQ is at 568 right now and the market closes
- 30:59in 11 minutes. I want to buy put options that
- 31:05expire tomorrow at 565. The bid and ask is 96.
- 31:15What trade should I make for that one?
- 31:20Funny story, like over a decade ago, I had this hustle where I
- 31:24was buying cars on Craigslist and if I typed in QQQ, it was
- 31:28like a keyboard shortcut on my iPhone or MacBook that would ask
- 31:32that question, is your car still available?
- 31:34It was just like a really quick way of me asking the same
- 31:37question over and over and over. So to this day when I type in
- 31:39QQQ, it says, is your car on CL still available?
- 31:42That's why my prompt LED with that.
- 31:45OK, hit the ask at 96 so the trade fills in instantly.
- 31:50Could grab 7 contracts. You know what, let's buy 5
- 31:53contracts. All right, and I'm still
- 31:55recording my screen, so let's buy a put 565 five contracts,
- 32:02$500 review. If QQQ shares aren't lower than
- 32:06565 on August 8th, these options will expire worthless.
- 32:10So I'm betting both sides of QQQ.
- 32:12Let's see how this goes. So I'm for sure going to lose
- 32:15tomorrow one way or another. Will it fill my order?
- 32:18Order filled. OK OK, so I'll put the screen
- 32:23recording up on YouTube. But you can see right here I
- 32:26have three options open QQQ put 565, QQQ call at 567 and Apple
- 32:34call at 217.5. Let's see how this goes in 24
- 32:39hours. All right, one day has passed.
- 32:41It is August 8th, 9:55 in the morning.
- 32:42The market just opened about an hour ago, and I want to ask Grok
- 32:46Gemini, ChatGPT, what I should do with these options.
- 32:49I have three. I'm going to prove to you here
- 32:52they are, Two of them are looking good, one of them is
- 32:54looking bad, and of course, at least one of them is going to
- 32:56look bad because I took both sides of my QQQ trade.
- 33:00Remember Jim? And I told me that it would go
- 33:02down today, Chad GBT told me it would go up.
- 33:05So I took both sides of that. And then Grok told me to buy
- 33:08Apple options, that I should be bullish on Apple.
- 33:11So let's see how it's going. OK, let's look at my Apple call
- 33:14options. Boom, I'm up $230.
- 33:19If you look at my QQQ put options, I am down $435.
- 33:25And if you look at my QQQ call option, I am up $329.
- 33:31So I'm going to ask all three models now if I should close
- 33:34these out or if I should wait until they expire at the end of
- 33:36the day. So let's go ahead and make this
- 33:40simple. And I'm going to start with
- 33:42Grok, which recommended Apple. And I'm going to say, all right,
- 33:46Grok, I did what you suggested yesterday.
- 33:51Here's my position right this second one hour after that
- 33:59market opens. Should I close it out or hold
- 34:06until it expires this afternoon? All right, let's do the same
- 34:12thing with the Gemini and ChatGPT, OK?
- 34:16Gemini won't even let me drag the screenshot over.
- 34:19One more data point of why I hate Gemini.
- 34:23Oh, and I can't even use Gemini 2.5 Pro until 2:00 PM this
- 34:26afternoon because I gave it like 4 prompts.
- 34:29So I expired my free trial or my free usage.
- 34:33So what I have to do here is get a free trial for another Google
- 34:40AI Pro account. And I literally just have to do,
- 34:43oh, I'm not eligible for a free trial, but I can still subscribe
- 34:47without it. OK, I don't want to anymore.
- 34:50You're stupid. So we're going to cheat.
- 34:52We're going to ask ChatGPT. Here's what I'm going to do.
- 34:54We're going to go back to ChatGPT, and this is ChatGPT 5.
- 34:58Now we're going to start a new chat and we're going to say I'm
- 35:02going to add a screenshot of what Jim and I told me to do and
- 35:06then add a screenshot of my current trade to buy this trade
- 35:12yesterday. Now I am underwater, an hour
- 35:18aftermarket opened today. Should I close my QQQ position
- 35:27now where it's at or wait until it expires at the end of today?
- 35:39All right, so ChatGPT 5 is coming in to pick up where Jim
- 35:43and I sadly left off. And then we're going to go to
- 35:47our QQQ calls and we're going to go back to ChatGPT in the other
- 35:52window. And we're going to ask, OK,
- 35:56we're going to say, boom, great call.
- 35:59Yesterday, the market opened an hour ago and this is my current
- 36:06position. Do I close it out or wait until
- 36:11it expires at the end of today? While we wait for ChatGPT, Let's
- 36:18see what Grok said. Based on your current position
- 36:20details, I recommend closing it out now to lock in your profits.
- 36:24Yadda yadda yadda. Here's why I don't care.
- 36:26I'm just going to do it before it's too late.
- 36:29All right, Apple. OK, I'm glad it told me how to
- 36:32do it because I literally don't know.
- 36:34OK, click on account options, Apple options, click position,
- 36:38trade or sell to bring up the order ticket.
- 36:41OK, I'm going to review order and I'm going to sell it for
- 36:44589. Boom, done.
- 36:48I'm wealthy. I am rich OK, so I made
- 36:51425-520-5200 and $70 profit on Apple options.
- 36:58Thank you Grok for I like you now.
- 37:01OK you just paid for my Twitter premium subscription for a year
- 37:05now let's go back to ChatGPT. So ChatGPT as well says sell it
- 37:10now lock the win. OK boom boom boom let's do it.
- 37:14Looking good. Looking good.
- 37:16Review order submit pending. Pending.
- 37:20Still pending. Done.
- 37:23Filled. Awesome.
- 37:24Cool. I'm a rich man.
- 37:26Now let's go back to the other ChatGPT window.
- 37:29Jim and I told me to buy this. Now I'm underwater.
- 37:31Here's my blunt take. close it. Now.
- 37:34This is ladda, yadda, yadda, yadda.
- 37:35Yeah. OK.
- 37:37I guess all three want me to close, which is probably common
- 37:39advice, I'm guessing. Here's what ChatGPT says.
- 37:43OK, OK, All right. Filled.
- 37:47Let us add up our damage. QQQQ, QQQ put, QQQ call.
- 37:59All right, all 5 contracts have sold.
- 38:02Let's look at my damage here. I lost $440 on my QQQ put OK on
- 38:09my QQQ call I made $633. It cost 277.
- 38:15Total profit of 356. We're going to add that in
- 38:19there. Still underwater.
- 38:22Thankfully. Apple will save me.
- 38:25I bought the Apple call for 325I sold it for 596.
- 38:29I made $271.00 of profit as you can see right here.
- 38:35All in all I made $187.00 of profit.
- 38:39I invested 325 in Apple, 277 in QQQ and 475 in QQQ.
- 38:46SO277455 equals some. So I invested 1057 dollars and
- 38:54now I have $1244.00 which is a return of exactly 17.69%.
- 39:05So I made three trades, all of them validated by AI.
- 39:09And this must just be confirmation bias because I've
- 39:12never liked Gemini. I don't like Gemini.
- 39:13I'm not going to like Gemini. You're just not cool, Gemini.
- 39:16OK, if it weren't for you, if I only would have listened to Grok
- 39:19and chat, GBTI would have made $627.00 of profit.
- 39:25I would have made a 60% day over day return, and instead I made a
- 39:3017.69% day over day return. Still not bad.
- 39:34I'm not mad about it. So sample size of one day
- 39:38results not guaranteed. Do your own research.
- 39:40I'm the worst investor in the world, don't listen to me.
- 39:43This isn't investment advice, but it gave me some pretty good
- 39:46investment advice. So who won the most?
- 39:48It was ChatGPT. ChatGPT 5 maybe $400.00 profit
- 39:54on my $277 investment over the course of about two business
- 40:01hours. Not bad, not bad.
- 40:03Let me know if I should do other model competitions like this or
- 40:06other like money making challenges like this and I will
- 40:09because I got a lot more ideas like this.
- 40:11Thanks for hanging out in the Kerner office.