I Asked AI To Make Me As Much Money As Possible In 10 Days (RAW RESULTS)

0:00OpenAI's frontier model, GPT-6, Astra is out and people are claiming this model is so intelligent it might be the start of AGI, artificial general intelligence. And I mean, that's cool and all, but can it actually make you any money? So, for the sake of humanity, I'm putting it to the test and for the next 10 days I'm giving GPT-6 Astra full control of this laptop and letting it run a real business. It's going to research products, make them, and list them for sale 24 hours a day. And by the end we're going to see exactly how much money it made. But first, we need to set

0:33some ground rules. Rule one, AI does the work. Every product has to be researched, created, and listed by Astra. Two, I can only check up on it once per day. Once per day I'm allowed to see how things are going and give it some direction only if it needs it. And rule three, probably the most important rule, I'm not allowed to post this [music] anywhere and get my audience to buy from the business. If Astra makes any money, it has to come from real customers finding the business themselves. So, you're probably wondering what business am I actually going to make an AI run? Well, I need

1:07something Astra can operate almost entirely from this laptop. So, that rules out anything physical. What's the opposite of physical? Digital. Yeah, baby. We selling digital products. And since I'm not allowed to tell anybody about this, I need somewhere customers already hang out, which brings us to, drum roll, Etsy. With over 86 million active buyers and the ability to sell digital downloads, this is the perfect place for Astra to set up shop. So, I currently got Codex taking control of my computer and creating Etsy listings automatically. [music] Let's see if we can make some money. So, first job for our mega mind AI, come up with a name. And this is where we hit

1:49our very first bottleneck, because turns out asking AI for a cute Etsy name isn't exactly an original business strategy. Eventually, we exhausted ChatGPT's greatest hits and landed on Little Wonder Studio [music] AU. How charming. Then, I gave Astra its next job. Make the store look slightly less like it was started 5 minutes ago. It went ahead and designed the logo and the banner and before I knew it, the shop front was complete. It went through this cozy little Australian art shop and it even generated the shop's bio. So, we officially have a brand. Now, all we're

2:25missing is something to actually sell. Okay, my business model is pretty simple. I get Astra to make the art. Then, I get Astra to list the art on Etsy, all right? And then someone buys it, right? Someone buys the art and then and then they they download it, right? They they download the piece of art. What ends up happening is they turn it into a Christmas card, right? And and [music] then and then they end up they end up I only make it once. I only have to make it once, okay? That's the big thing. I only make it once and then and then it's just infinite money glitch

3:01because Etsy sells it again and again and again. It's literally just infinite money, right? Simple. We're going to make money with this thing. So, again and again and again and again

3:17Okay, so we've just gotten our first Etsy listing and I'm actually pretty amazed with the result. However, I have realized that it is taking up a lot of our weekly token usage allowance. And so, in order for this to actually be efficient and a sustainable business model, [music] we need Codex to be more efficient. So, basically, Astra was making Etsy listings, fine. Oh, no, that wasn't the problem. It was the fact that it took 4% of our weekly usage limits to get a single listing live. At this rate, even in perfect conditions, 100% of my weekly usage limits would only get me 25 listings. Now, that might sound like a

4:07lot, but it's really not. 25 listings a week is only about four a day. At 10 minutes a listing, that's 40 minutes of work and 23 hours of Astra doing absolutely nothing. And on Etsy, every listing is a point of discovery. If I didn't fix this fast, then this business plan was going to fail. And so, I've

4:28actually sent Codex another objective. And this time, I've asked Codex to create another three listings. And after every listing, I want you to try [music] and optimize uh yourself, so that you are using less tokens each round. And hopefully, we can get down to just 1% or even half a percent per listing. So, I spent the next few listings figuring out how to optimize Astra's workflow, [music] and lo and behold. Okay, wow, this is crazy. Codex has just created its first listing of the three listings we told it to create while optimizing its workflow, and it already looks like it's down to just 1 to 2% just from that very

5:11first listing that it made. This is looking very promising, guys. Now, at this point, you're probably wondering why I'm taking this so seriously. You see, I've actually done this before. Three years ago, I ran an Etsy shop that brought in over $10,000. [music] I'd put over 6 months worth of my weekends into it, and it was consistently making me $600 to $1,000 a month. I was in my early 20s at the time, and I was working at a cafe, and I'd been working at that same place for over 5 years. And I started thinking, [music] if I can keep growing this thing, I might actually be able to leave. My

5:46friends and family were so impressed, I ended up becoming the Etsy guy, and I was starting to believe I could actually do this for a living. Except, there was one problem. In my Etsy store, I was selling Harry Potter stickers, and I didn't have the commercial rights to sell Harry Potter products. So, without

6:03any notice, Etsy just banned my store. Over 6 months of work gone in an instant. I'd started imagining leaving that cafe, and now I had to start all over again. I tried another store, but I never got anywhere near the same amount of success. So, I'm giving Etsy another go, but this time I've got a GigaBrain to help me. Also, if you're enjoying this experiment, give the video a like and subscribe if you want to see what I make AI do next. All right, it is officially day two, and I want to go over a couple extra things that I did yesterday that I didn't get on camera

6:38before we get into the actual stats of the store. Did we get any sales? Let's have a look. But, before we do look at that, we end up finishing with a total of 18 listings uh yesterday. I could have gone harder, but I didn't want Etsy to flag us since we are a new store. I don't want them like flag me off as a bot or something

6:57like that, even though we are a bot. But, yeah, overall, I made some further optimizations. So, to start yesterday, our very first listing that we generated, it cost us a total of 4% of our Codex usage, right? And so, the whole goal of yesterday was to optimize and see how low can we get the usage limit. And after going back and forth a bunch with Astra, we managed to get it down to drum roll

7:270.33%. Yes, heard that right. From 4% of our weekly usage down to just 0.33% of our usage limit per listing, okay? So, after every three listings, GPT-6 Astra researches for a new theme or topic, and that includes writing up the description, generating the image, everything is packed neatly into this 0.33% So, we've gotten super lean now. That means we went from a total of 25 listings with 100% of our usage limit to now being able to do over 300 listings. Like, that is incredible. That is exactly the the goal, and that was exactly the amount I was trying to achieve. 300 listings. If we don't get a

8:21sale after 300 [music] listings, then we're doing something wrong. Speaking of, let's have a look at our stats. So, okay. Not bad. In 1 day, we got a total of 14 views and seven visits. Unfortunately, we haven't got any orders yet. Someone actually has to buy it. So, on day two, we started to focus on customer acquisition. Today, I actually want to introduce a new strategy into the workflow known as eRank. Now, if you don't know what eRank is, it's basically a keyword research tool that lets you look at certain keywords on Etsy and see how much competition they have. So, for

9:04example, we're selling uh PNG downloads. So, if we can look up PNG right now, we are able to see the search volume. And if you scroll down, you can see the competition. So, Christmas PNG, there is 842,000 listings fighting over 11,000 searches a month. Now, with this tool, I'm hoping Codex is actually going to have a much better chance of creating listings that people actually care about. And so, we're basically just going to looking to go

9:39through this. Mahjong PNG has 1,000 search results for just 642 listings. So, this looks like a really good opportunity to create Etsy listings around Mahjong PNG. So, I'm looking to just keep this open on the browser and give Codex access to this tool and hopefully it doesn't increase our usage by too much. And now that we've got our optimizations set in place, I think we can just let

10:08Codex continuously run 24/7. So, now what I'm going to do is I'm literally just going to prompt it and be like, "Yo, I want you [music] to start using eRank to keyword research and see [music] and find good keywords that get good search volume. [music] And the goal today is to generate maybe three listings every hour. [music]

10:54And that's the prompt for today. You can see here it's already started generating Mahjong style, golf style, and thrifting style listings, which is exactly what the research showed. These look beautiful, by the way. [music] So, by the end, we should have a total of 72 listings this time tomorrow, which is crazy when you think about it. So, that would put us at a total of 90 total listings in our first 2 days of launching our Etsy store. That [music] is ridiculous. There's just no way that one person running their [music] own Etsy store would be able to produce that many listings by themselves without the help of AI. Like, the competition

11:43is crazy. You are now so competitive in this industry using AI. I'm honestly I'm pretty excited. I feel like these upgrades to the workflow are actually going to greatly improve and increase our chances of getting that first sale. And I think that first sale is really important because once you get that sale and you make that customer happy, if they leave a good review, then that's really the only thing stopping us right now is we don't have that five-star review next to our name right now. So, people can feel a bit sketched out. But if we can get that review, five-star review, then we're basically on the same boat as all the other

12:25sellers on Etsy. We should be way more competitive. I'll let these run and I'll check back in with you guys tomorrow. Peace. Still going, babe. Still going, man. All right, so it is currently the next day. [music] Let's have a look at the stats. As you can see, Codex have been going hard with

12:50a bunch of listings. We've set up automation, so it's doing multiple listing [music] every single hour. If we have a look at the actual stats, we can see a total of 53 listings in the shop now, which is crazy, by the way. 53 listings in two days is ridiculous. Now, before I show you the stats, we do have some very exciting news to share, okay? We got our very first message from

13:22an actual customer. This guy, Quang. Thank you, Quang. I really appreciate it. He said, "PSD file includes [music] layers." Now, I'm not entirely sure what that even means, but I know that we don't really have layers, so let's shoot him a message. I'm just going to say no layers. But, it's cool to see that we're getting interest, you know? Like, it's only been 2 days [music] and we've already got someone messaging us about one of our products. Now, it wasn't an order, but someone had found our tiny shop and had enough to ask a question. Suddenly, this all felt very possible. And if you're watching this thinking I'd love to try

14:04something with AI, but I wouldn't know where to start, I've put together a community called AI Solopreneur Academy over on School. [music] I show you how to take an idea, build an actual app with AI, and get it in front of people, even if you've never written a line of code. You get the tools, prompts, and custom skills I use, breakdowns of what I'm building, and weekly live calls where you can bring your own project,

14:26and I get to roast you. So, if you've had an idea sitting in your notes app for 6 months, come build it with us. The link's in the description. [music] Let's get back to the video. So, we have a look at the stats, we can see 37 total views, 18 visits, which I'm pretty happy with. I'm not going to lie, that's pretty that's pretty good. So, we can see here what our most viewed listings are. We've got

14:51the rainy day umbrella, ballet shoes. So, today is actually a really important day because we are now going to be implementing a new strategy onto our workflow. Now, I'm not too sure if you guys have noticed, every single store on Etsy does the exact same tactics. They've got bundles, and they've got sales. And this drives conversions because you're getting more bang for your buck, and you're also creating this sort of scarcity because if you click on it, you're going to see 60% off, sale ends in 2 days. Like, oh I I better buy this. I better buy this quick. So, what I want to introduce to

15:35Codex now is bundling themes together. I basically want Codex to experiment and try doing sales ranging from 10% all the way up to 90% and five trying to find the sweet spot. All right. I've just sent a really big prompt to GPT-6 Astra basically explaining the new strategy. Let's see how it goes. But, I actually do think that this new strategy is actually going to greatly improve the

16:07conversion rates on our Etsy store. Because up until this point, we may have taken the blindfolds off and been able to see what's trending on Etsy, but while everyone's using a crossbow, we're using a bow and [music] arrow. So, by being able to implement sales and bundles into our Etsy strategy, I really do think we're going to have quite a competitive advantage considering that GPT-6 Astra never sleeps. And once we have all of this configured, I'm actually going to bump up the listings and I'm going to try and get GPT Astra

16:44to do 100 listings in the next 24 hours. So, we made the offer bigger, more designs, bundles, discounts. I was convinced we were getting closer. At this point, I decided to let GPT-6 Astra run over the weekend and check back in on Monday. All right, guys. So, it's not looking too good. I'm not going to lie. I I let Codex run over the weekend and we've done

17:18a bunch of listings now. However, I'm not going to lie. It is a little bit demotivating, man. Like, if we actually have a look at the stats for the store, um you know, especially like yesterday, we only got like four total views, three visits. Today, we haven't gotten any views, so yeah, I don't know, man. I'm going to keep trying and we're going to turn this around, but it's not

17:52looking too good. 187 listings. I kept thinking the next batch [music] would be the one, but every time I checked, I was still waiting for that very first order. I'd already had one attempt at rebuilding an Etsy shop that went nowhere. And looking at [music] that dashboard, I was starting to wonder whether it would happen again. But a warrior is not defined by how many [music] battles he wins, but by how many times he rises after

18:20defeat. Most surrender at the [music] ounce of difficulty. Yet those who succeed understand something the others [music] do not. Defeat is only temporary. And as long as you continue to rise, to learn, [music] and to find another way forward, you are yet to be truly defeated. For the only true failure is to stop trying. So, I put my head down and locked in. I kept looking for one more thing to optimize. Better keywords, bigger bundles, smarter discounts, anything that might get us that first sale. Then,

18:55I checked the shop again. We just got our first sale. We made a whopping $1.16, okay? Now, it might not sound like a lot, but this is probably the most valuable $1 I've ever made in my life. She bought a collection of these cute little piglet Halloween piglet listings, which honestly I don't blame her. They are adorable. Who wouldn't want to buy a pair of those, right? I feel like a a new person now. Now that I can see that the first sale is possible, I'm excited to venture on and try and get a couple more sales before the

19:40challenge is complete. [music] But one sale could still be a fluke. Could we get any more before time runs out? So, for my last check-in, I had Astra go back through the shop, tidy up the listings, and try a few more discounts. We've finally got one customer. I wanted to give us the best

19:57chance of getting more. [music]

20:06[music] And so, on the last day, I checked the shop one final time. I can't believe I'm saying this, but we got another sale. It is our last day, and we literally got another sale. Janelli, one order for $3.82. Oh! I'm so stoked. I'm so stoked, man. And the second order was for this little Halloween sheep bundle. Someone looked at these sheep in their tiny Halloween outfits and got their wallet out. Fair enough. They are

20:51pretty cute. Dude, I feel so good cuz it's like it really does feel like the shop is starting to gain a bit of traction now, and you know, considering where I was a few days ago, I I genuinely didn't think I was going to get any sales. So, the fact that in such a short amount of time we've already gotten two sales, like holy sh-

21:17man. I actually can't believe this because on the last day, too. Like, [music] what are the chances of that? Well, let's have a look at the numbers, I guess. So, the final numbers, 191 listings, 48 visits, and two orders. Bringing us to a grand total of $4.55 in revenue. That's before Etsy's fees and the cost of running this thing. We also had eight people put them into baskets, which were eventually abandoned. So, I would have loved to turn those into additional orders by adding discounts for abandoned baskets, but yeah, they weren't. So, yeah, that

21:58was definitely a missed opportunity. Honestly, the thing that surprised me the most about this whole experiment was how quickly I stopped being impressed by Astra. On day one, watching it make a logo felt ridiculous. By the weekend, it made a whole shop full of products, and I was sitting there thinking, "Okay, but is anyone going to buy them?" I'd got so caught up in how much it could make, I started treating the numbers of listings as the scoreboard. But, I still needed to give someone a reason to buy. Astra did the research, made the artwork, and put the listings together. I still gave it direction, checked the results, and

22:32changed the strategy. And by the end, it had helped me get two real orders. Now, I do plan to keep working on this, so like the video if you want a part two. However, I'd need a lot more time and a proper look at the cost before calling this a business that pays for itself. And coming back to Etsy after what happened with my old shop, I'm glad I gave it another go. I didn't get my $10,000 back, but when that first order came in, I got a little bit of that feeling back. Someone had found our shop

23:01and wanted something we were selling. So, can AI make me any money? Well, yeah. In 10 days, it made me $4.55.