Speaker 1I've used Gemini for Argon non-stop for the last seven days and I had to come on and share what I found. You know, most people are going to use Argon just like every other AI, but there are three specific problems where if you just point Argon to it, it can help you build a real business and no one on the internet is talking about it and no one's telling you how to do it. I'm going to tell you. By the end of this episode, you're going to know how to use Argon three ways. The first is Vibe Marketing. How to use Argon to get customers to your startup. I truly believe it's the best AI for writing right now and I'm going to show you how to set it up to write just like you and all the skills and everything. Second, we're going to talk about Vibe Manufacturing. Argon is really good at helping you design personalized physical products. I'm going to show you exactly how to do it, all the CAD files and everything. And third, how can you use Argon to point it at legacy SaaS products that people are still paying for? And just build a better version. I'm going to show you that. I'm going to give away some real sauce around a technique for actually getting customers there. I'm going to walk you through how I would do it. And I want to give a quick thanks to Google for sponsoring today's episode and giving me early access to Argon. Everything you're about to see, I actually built and these are all my personal opinions. And if you stick around to the end, I'll show you six hacks to get way more out of Gemini for Argon than most people will figure out for months. So let's get into it. And I'll see you at the end. Okay, things to know about Google Argon. What is it amazing at? Well, it's amazing at marketing with AI. So that's important. We need a tool that actually brings you customers. I'm going to show you how to do that and where it's good at it. It's good at physical AI, aka Vibe Manufacturing. So how do you take STL files and send it to Bamboo and 3D print it? I'm going to show you how to do that on demand using Argon. And it's good at complex engineering. So I think that you can point it at, you know, very legacy SaaS products, add your own spin to it, pick a niche, go even nicher, and it could rebuild it relatively quickly. And it's, this isn't vibe coded software. It actually, it is tested. It feels really good. And I've just been so impressed with how you can actually point to complex engineering. I will, I want to get this out of the way. Anti-gravity is not my favorite harness. I think that Google has some catching up to do with anti-gravity. And the more I played with it, the more I was missing some of the other harnesses. Now you could still use Argon in your cursor or in other harnesses, open code and stuff like that. I just wish that anti-gravity would catch up. And I think it will catch up. And I think that, you know, one reason I think people should actually pay attention to Argon is Google has so much data and they're making so much progress. And now they're at the frontier level in terms of a model. So let's talk about marketing with AI. I first want to talk about a website me and my team built in 2021. This website took 40,000 hours to build. And it's a, a website that was built for my holding company. And it just has all the different buildings for all our different companies here. You see LCA, our design agency, Boring Marketing, which is our marketing with agents company, careers, all this stuff, right? I can't tell you how much money this costs, but I can tell you that, you know, the business has probably brought in more than $40 million of revenue. And a large part of it has to do with this, has to do with this website. And people always tell us, how good their website is. I asked Argon to recreate the website. Or actually, I just said, hey, I want a website like this, but can you make it better, basically? And it created this. And it's, first of all, beautiful. I actually gave it the inspiration for, I was like, make it more Mondrian style, Mondrian, the artist. And I think it did such a good job. You know, when you compare it versus the one, that probably cost us $200,000 to build. I mean, yes, this is really nice, but it's sad in the sense that there's no cars, there's no movement. And the web, the internet has moved, right? And you want things that are more, you know, that you can interact with. And you can see here, there's this like world passport. When you, you know, click on the buildings, there's some sounds and like it gamifies going, checking out all the different, you know, things that are going on. And you want people to do when you want to go to their website, right? And I just think like fun little games and things that you can do. Look, shoot a three pointer. Boom, boom, boom. This is crazy. It is crazy that you're able to create this. And when I say marketing with AI or vibe marketing, what I mean by this is, why are you creating a normal website in an Argonne age? You know, it doesn't make sense anymore, right? Because you want to stand out. And if you could, you know, how do you stand out? Well, the limit isn't the model anymore. The limit is now the imagination. Because when you have a model this good, with an engine this good, then you can create stuff like this. Now, my first prompt looked like this. So, and I first was like, hey, make a better version of LateCheckout.studio and made this. And I think we can all agree that this is pretty brutal. This is really brutal. You know, it got the concept, but it really felt like, you know, when Mark Zuckerberg announced the metaverse and it was like that, that like Zuck in 3D and everyone made fun of him. We'll put that photo up there. It felt like that. And I think the question in this whole AI era is like, how do you, how do you go from this all the way to this, right? What is the way to do that? And one thing I learned about Argonne specifically is it needs good references for good creative. I don't believe that you can get great creative out of Argonne unless you give it references. So it actually, in some ways, it's kind of great because you don't want to outsource all your creative thinking to AI because then you end up creating stuff that doesn't really stand out. I think this has the power to it, but you do need to like, you know, give it, give it that, those references. Initially it created like all these, you know, way too many buildings and the buildings were everywhere. And I was just like, it's too much stuff. And you just have to like have the creative eye to be like, it's too much stuff. Another, another example, was like the initial version felt like Roblox-y, like it didn't feel premium. And we weren't there yet. All that to say is this, this was a success. And one of the reasons why, frankly, I don't update this website is it's really hard to update. It takes a lot of time, energy and stuff like that. Whereas this, now I can just use Argonne to update it. I saw a really cool example of someone using it. Justine Moore, A16Z said, I asked Gemini for Argonne for an interactive museum of Apple computers and didn't provide any images or data. It scraped archival Mac photos, wrote a tool to clean the images, pulled 50 years of specs and built a 3D cover flow with UI that morphs to match the decade you're in. So when you look at this, look how good this is. There's just so much opportunity to whatever it is you're building, not to build something basic. And that's what I mean by marketing with AI is that this would have cost you thousands and thousands of dollars before, but now you just, you know, you can just go and create this. And this is the sort of thing that goes viral on Twitter. It goes viral on TikTok. It goes viral on Instagram. And then you can decide, by the way, what business you want to create on top of this if you don't already have a business, right? You can be like, oh, you know, I want to go and I've got this audience now of people into Apple products. Well, what can I sell those people? Is it services? Is it memorabilia? And we'll talk about physical AI and what are the things that, you know, you can do later with Argonne there. But I think that's a really one big thing I learned that I wanted to share with you all. The second thing I learned is, Argonne is really good at writing. So it's extremely good at it. And I'm going to let you in on a little tip here, how to make Argonne an incredible writer in one step. So you say, I want you to research everything about how I write. This is what I wrote. My name is Greg Eisenberg. I write on X. I do email newsletters, how I speak on YouTube, everything about me. I want you to internalize all that stuff into a skill. Once you've done that, I'm going to test you on it. I'm going to give you a topic and I want you to write an essay on it and we'll see how you do. And what it does is it basically creates the skills and what I should have done is actually asked it to do slash subagents to send subagents to go and do that. So you can just go slash subagents on anti-gravity to go and do that. But what it does is it created a skill package. So it created a skill MD file with the core voice operating system, sentence rhythm rules, mental models, the structure I use is it created a voice and lexicon MD, like signature vocabulary, things I've coined. And it's got all this stuff about me, which is crazy. Verbatim samples and templates MD, plus all the structural templates. This is something that if I was hiring like a head of content, they would be able to do. And they did it perfectly. It internalized exactly how I write and speak. And this is just so cool because I find that one thing that AI isn't that good at is writing. It's crazy because it's so good at other things, but I don't know why it hasn't been such a good writer. And Argonne to me, at this point, is probably the best frontier model for writing. If you look here, I basically said, write me a post on marketing engineers. And it wrote me a post and it was amazing. It was like, you know, exactly, it was actually crazy. Because it was exactly how I would write. The old world versus the new world. You know, talking about micro products instead of PDF lead magnets. Things that I've said in the past. Tweet-sized landing pages. You know, I have this belief that launch your landing page initially not more than 280 characters and stuff like that. It had all this stuff. Now, if you don't have content online about you, like me, just literally upload it to your anti-gravity and have it learn about you. Like, you've written stuff, you've written stuff before, right? So you can have it do this. I really like this 95-5 rule. AI does the plumbing, humans bring the taste. And what's really cool is it even signed off, I signed off my newsletter at gregisenberg.com with Be Well, Greg, and it had that. So if I hired someone to write this newsletter, I would have been like 9.2 on 10, probably. I would have been stoked. I would have been like telling my co-founder, hey, you're not going to believe this writer I found. And so I think that from a vibe marketing perspective, marketing with AI perspective, using this to write more should be on your list of things to do. I don't know what you're using to write if you're just doing just humans or if you're doing some mix of Astra or Opus. But in terms of what I'm seeing as a writer, as someone who puts out a lot of content there, this is definitely something that I think is quite good. Let's move on to physical AI. Now a lot of you have been asking me about vibe manufacturing. I had this tweet that went viral around vibe manufacturing and how this is the age to actually go and create these files called STL files, send them to 3D printers, and build businesses around customized physical products. But how can you do it? Well, today I'm going to show you how. Argonne is actually amazing at it. I actually asked Argonne to build me this vibe manufacturing studio. So this is a big idea here. I hope people take. So my belief is that you could be able to customize physical products. I think it's going to be a bigger and bigger thing. I think the idea is there's millions of Shopify stores, right? What if you can embed this widget that allows people to actually customize products on Shopify stores? And they could buy it for, I don't know, $39, right? So you have here, you can see how long it's going to take, how much it's going to cost for the weight, how much is bamboo, print time is going to be. And by the way, if you don't know what bamboo is, it's a 3D printer. I think it's a Chinese company. What suggested retail would be one printer overnight that takes eight hours for four units. I think this idea around, basically embedding physical AI customizers on Shopify is an extremely huge idea. Now you might be asking, okay, what would you actually create? Well, think about, let's talk about the, I don't know, your snowboard company. I know Toby from Shopify started with a snowboard company. It was his first Shopify store. Maybe you create, you know, famous hills, famous mountains, and you allow people to customize it. Maybe you create, you know, famous snowboarders and allow people to customize it. Maybe you could put your own face and your own body in there and you allow people to customize it. When people customize things, they feel more attachment to it. And GIFs in general, by the way, physical GIFs and these niche SKUs, I just see seeing more, more and more of a thing. And, you know, one example here is an Espresso Portafilter station. You know, I would pay for this. And you can see the demand signal, it goes through Reddit. It goes through the subreddit r slash Espresso who spend $2,500 on machines and gladly pay $49 for custom Portafilter stations. Just makes total sense. And I think this is like a big business. Argon nailed this. You can see here, you know, when initially I asked it to create a wireframe, show the layers, and then show the CAD. Like when I see this, it's almost like an indescribable magical feeling that you're able to create this. And there are so few people doing physical AI. So worth playing with. And I was. impressed with how well Argon succeeded here. One of the cool things you can do is see this export ready to print STL file? Well, that STL file can send to your Bamboo printer if you own it. If you don't own one, you can do SLAN 3D print on demand API. So I think a lot of people are going to be like, oh, I don't want to own like 20 3D printers in my house or my office. Okay, don't. Just use an API. Okay, and have it print there. You literally, all you have to do is aggregate demand, come up with an idea that folks are going to resonate with. But you know your niche better than anyone else, right? So you have that insight. And then you go into Argon and you basically say, like I basically was like, hey, just build this embeddable Shopify widget just so I can show people. It was literally a one prompt. And then start testing it. See if your audience likes it. Get feedback. What do they wish you can have? But I think this idea around niche SKUs, physical AI, it's just getting started. How I would actually get to work on this. So you have Argon build the product customizer. You embed in your store. So the customer picks like a size, a color. You have a live 3D preview so they can see what's going on. The customer pays on charge. And Shopify, I just start with Shopify. They're the number one, right? Of course, you could do it in the future in other places. You check the custom print file. So you've tested size limits. And then you print and ship the product. So it could be your printer or a partner. By the way, you can also just sell this as a service to other Shopify's. So you can actually just set up a business. Like here's a business idea for you. Like just go. You can go and do this for other companies. You go to them and say like, I could help build more fandom for your audience by having physical products so it's on their desk. They see it every day. It's when they walk into their homes. They see it greeting them. And they're going to be thinking of your brand. And you're going to make 50% margin, 40% margin, 70% margin depending on how you price it. Is that something you'd be interested in? And maybe you split the margin with them. Make it a no. It's a no-brainer offer. These are things that you can now do. So the third way I want to talk about where Argon really excels is building SaaS products. So boring SaaS products. Now, I already hear the comment section. People are going to say, well, Greg, SaaS is dead. What is the point? Well, I take a deep breath for that one because I see that on X all the time. People saying SaaS is dead. AI can do SaaS. Therefore, it's over. Those are indie hackers selling to indie hackers. But the vast majority of people are fine to pay for software that helps make their lives easier. Now, here's the way I would approach building a SaaS product. And I think it's different from how other people are talking about it. And then I'll show you how Argon goes and does it. So first, go and find a recurring, ideally boring, problem. You can find these problems on g2.com, Captera, Idea Browser, something I built. By the way, it's literally free to sign up to get a daily idea. And then from there, I would go and build a very focused problem. Sorry, I would build a very focused product. You can be looking on g2 and these other places, even Reddit, to go and see what are people complaining about and then obviously put that into your product to make something better. There's so many legacy pieces of software that are just their legacy. They're old. They haven't been updated in so long. So there's an opportunity now to rebuild something that really fits the needs of people that maybe is actually a little bit more niche in a good way. And then you go and make a useful free tool. A lot of people go out when they think about building SaaS, they think, oh, I'm going to go and build something, charge $50 a month, and I'm going to get 10,000 customers. I understand why that's the approach, but I actually think that getting a free tool out there is such a good way to get it in front of likely buyers that make your second or third or fourth tool paid and just try to think about what's one tool that can go and attract a lot of people. And then from there, you can go and essentially create a product studio around that ICP. And your studio could be powered by Argon because Argon has become such a good way to actually build software that actually works. And when I say that is we've all vibe-coded stuff, but when it comes down to deployment, it hasn't been good or it hasn't been secure. And for my testing, anti-gravity has done quite a good job of that. So I'll give you, an example, I gave anti-gravity this prompt, build a working web app for a commercial cleaning company that turns site visits notes into a reviewed price proposal. This is a focus replacement for quoting spreadsheet. The main experience should be creating a quote. And then you give it customer complaints, source links, anonymized workflow here. And if they don't have it, here this would be your G2, this would be your idea browser. But if you don't have it, you can actually just say, go and find it. And it should be able to find something. And then find and inspect the project, explain your implementation plan briefly, then build it using the existing stack where possible. Then you just write it all out clearly. You say don't deploy it, send proposals or anything, and then give an honest list of remains unfinished. This format is the format for building these micro-SASes. It helps get you in the door with a lot of these companies. When you look at what was built, let me pull it up actually, it's basically these interactive margin calculators. When you look at what was built, I mean, again, this is not the most beautiful thing. And that's, like I said, if you want to get beauty out of Argon, you really need to give it references. And you've got to work with it to get it to that point. It has the capacity to do it, it just, for whatever reason, in my experience, hasn't gotten it on prompt one. But where it nails it on prompt one is on the engineering. And all these little details that you can see here, it does such a good job at. I mean, I'm not going to bore you with all the boring parts of the workflow, but this format, in terms of these quotes and stuff like that, I think is a big idea. One that I can see people doing. And if you're going and you're competing against someone that charges money for this, and then all of a sudden you're like, hey, this is free. No one talks about that. No one talks that there's thousands of these boring SAS products that charge money, then now you can go and sell it for free. Literally, zero dollars. In the past, I would be scared about vibe coding something, but we've reached this point with Frontier AI that you can build something like this. And it works, and it's probably even better than what the legacy products have. So the question becomes then, what do you do when you sell a free, a paid thing for free? Once you've built the trust, eventually you sell them something. So totally was impressed with how Argon performed here and happily surprised. So having played with Argon for about a week, I found out that there's six techniques worth using to get the most out of it. This isn't going to be rocket science to people who use AI a lot, but I will say it's going to help. So the first is, the second is plan. Slash plan. Just agree on what to build before it starts changing files. So for example, with my latecheckout.studio example, I should have used plan mode. If I would have used plan mode, it would have been a lot better. My first version would have been a lot better. The second is subagents. You can do slash subagents. That splits independent jobs, so research and checks can run in parallel. So the example I gave, around the marketing agents, creating the different skill files for writing better. In subagents, I would have got better output and it would have been less prompting and stuff like that. So subagents, you might as well use it. And it's basically this concept that you're not giving a task to one thing and it has to go into multiple different directions. You're just basically saying, hey, you're delegating. It's just literally delegating to different agents to do different things. Slash goal. One of the reasons why the latecheckout.studio ended up becoming really good was, I did slash goal halfway through it. I was like, all right, slash goal, make it amazing. I gave it a reference image. And yes, it does take more tokens, but I noticed that, output is a lot better. So for me, it is definitely worth it. And you have to think about the value for you. Think about how much we made so many tens of millions of dollars from that website. An extra several dollars in tokens is worth it, if that's even what it costs. Slash browser, you can actually have it open the app and test the actual user journey. How cool is that? Good thing to know. Number five, slash learn. So you can turn your corrections into reusable project rules or skill. You might as well do that. So you're going to have things that you're going to mess up on and you're going to want to have a better system, more skills, more project rules. You're only going to get better output, so you might as well use learn. And the last thing is, initially I was on Medium. I turned it off. I set it up to high. High has been my base on this. I've been enjoying using high. I do find the output's just a lot better in general. So I would say use high, personally, as the default. But start on Medium so you understand how to use the thing. These are just six techniques to get the most out of Argon. The biggest one for me being plan, sub-agents, goal. I mean, they're all kind of big. I mean, this is why I put them in here. These are some of the biggest ways to get the most out of Argon. So there you have it, folks. Gemini 4 Argon is out. Some of the things you can do. I think we're entering this new era of frontier model where these models are just getting so darn good at doing complex problems. I was impressed. I was impressed. I will say I was using Google AI Ultra, which is $100 a month. That gets you full access to anti-gravity app at the CLI and the up to one million output tokens. And thanks again for Google for hooking that up. The 1999 Google AI Pro membership does not include the anti-gravity app in CLI or Google AI Studio access. You don't get that. You don't get the one million output tokens. So that's just more for like text generation or web search, little tasks. But if you're looking to build, the Ultra is what you're going to want. And what can I say? I think I'll end with the limit is really your imagination. And the hardest part sometimes is just coming up with the ideas, finding the references, getting the inspiration. and finding a niche or ICP that you're going to fall in love with. I always say marry the niche, date the product. You can always try different products, but you're going to want to love the niche to serve them. Had a lot of fun playing with Argon. Interested to see if there'll be updates to Anti-Gravity. Hope that gets better soon. And I'm excited to see what happens in Argon or Gemini 4.1. So let me know what you think down below. I read every single comment. Thank you for coming along this journey with me. I'm here to serve you. I'm rooting for you. And I can't wait to see what you build.
PodcastTranscript AI