Jordan Metzner: This is all about bots. So get your bot going, my bot's ready, get your bot ready because Grok bot and every other bot's about to bot you up. Built This Week, breaking it down. Built This Week, we show you how. A fresh idea, a clever tweak, you locked in.Sam Nadler: Hey, everyone, and welcome to Bot This Week, the podcast where we share what we're building, how we're building it, and what it means for the world of AI and startups. I'm Sam Nadler, cofounder here at Rise Labs, and each and every week, I'm joined by my friend, cohost, and business partner, Jordan Metzner. How you doing today, Jordan?Jordan Metzner: Yo. Bot This Week. I love it. That's a new name for the podcast. We should rename everything.This episode is all about the bots. Bots are changing my life. I'm so excited to chat with you about bots. Just gotten pushed notification from my bot just now. So it's about time.Sam Nadler: Alright. Well, so, yeah, before... We're gonna jump into some custom bots you've built this week. We don't have a guest, so it's a it's a fun moment in time where we can jump into what we're building personally. But before we do, please don't forget to like and subscribe.I don't know the exact count, but I think we're approaching 50,000 subscribers on YouTube. Very, very close. New episodes out every Friday. And with that, let's jump into the topic of the week, bots on bots on bots. By bots, what are we referring to, Jordan?Give us the high level how this is different.Jordan Metzner: So, you know, I think, you know, everything's about agents and bots and there's too many names for different things. You know, obviously, we've got Open Claw and kind of the first iteration of bots and in that iteration. But I think we've seen a step forward and a step function change kind of with Muse Muse bots and Dot from OpenAI and then my favorite Crockbot. I think if you go back to like chatbot, you know, kind of the first iteration was like using ChatGPT to code, kind of pasting stuff in and getting it back. And then, you know, eventually we got like ChatGPT and other models like in the sidebar, and so they would go in and edit your code, and then, you we kinda went to, like, an abstract model, like the Cloud Code model, and you're just talking to it, and it's editing your code.You know, all of that is fine. And, you know, then we kinda got, like, I don't know if you've if you've used lately, we've gotten into, like, the browser use cases where, you know, ClodCode and and Codex and can now use your browser and use different tools in your browser and all that all that's cool. But I think the big step function is with GrokBot and Muse and some of these other ones is its ability that it has its own browser and its own computer. And it's using your accesses to then construct behaviors, workflows, and all these other types of things that probably weren't previously able to be done. So, you know, a simple one might be just, you know, I'm looking to buy a house in a certain area.I want you to search every day on the market. And, like, you know, maybe you would do that with Zillow, but, you know, maybe there's, five or 10 different websites in which there's like properties available, and so you might wanna do it that way. But what's really cool is you can connect it to a bunch of different tools. So I think you know as you know about this, but, you know, we have our email connected to Grockpot, which gives its ability to send emails or even prepare emails or reply to emails on our behalf as well as as well as connected to my LinkedIn. And so what's really cool is I can I can give it a command of say something like, you know, I wanna reach out to these certain types of people, and I wanna do this many people each day, and it'll go target those audiences and make those LinkedIn connections?And previously, you know, we used to have some automation software that used to do it. And if you think about even how that software worked, it was using your browser. Right? It would it would work, but it was always using, like, your browser to run these workflows. The difference here is it's using, you know, Grockbox computer per se.What I think is so interesting is as the beginning of a new agentic type of workflow where workers can say, like, you know, your job is to, you know, for sales, for example, go reach out to this type of audience, go get their information, go reach out to them, go customize the message, go test it, you know, and see what converts, try to make as many meetings as you can, respond to everyone in real time. You'll start to see this happen, like, over and over and over again. And it's so obvious that it's gonna come to the enterprise very quickly, kind of like the real actual copilots that are helping people do their jobs and do their work. And the more kind of tools that get connected to these bots or agents, the more effective they'll be. And so essentially things that were, you know, multiple step processes of workflows like, you know, respond back to that person, go check into the system, go have it to this, go have it to this.You know, you don't need to tie those together. You can simply, like, kinda resolve for what you're looking for and just... And let the the bot or the agent do its work. And so, yeah, it's been really fun to use, really fun to play with. Love to hear, you know, what you've liked about it.Sam Nadler: Yeah. I mean, I would love for you to walk our audience through a practical example. Personally, I'm about three weeks in into using Grock bot, and, you know, there's... You know, I feel like I'm just scratching the surface, but there's all... Already tons of, like, little daily actions that I've been able to offload onto my Grock bot.There's also some personal applications. I think it's the, you know, really obvious one is, you know, looking and booking travel. But, yeah, I would love for you to... I, you know, I know you prefer GrokBot, so maybe let's stay there. Why don't you walk us through, like, a a kind of a high level overview of one practical use case to kind of, like, put a little meat around what we're talking about?Jordan Metzner: I don't know if there's as much like show and tell because it's, you know, it is just the chatbot based experience, but, you know, I think you know the example of last week, you know, I wanted to stay in a specific hotel because it was across the street from the office where I was working, and the hotel's website said it was sold out. I had Grock Bot write an email to the hotel, so it found, like, who's the right person to find the email. It write... It generated the email. It wrote to the hotel.It asked for... See if we could get a room and if I could get a discount. And I don't know, maybe fifteen minutes later, someone from the hotel called me up, offered me 33% off of the room rate, and then I was able to extend that, you know, over to another guest and then, you know, basically save on a on one night one night stay about $200. So, you know, that was pretty light lifting. And I do think that these agents are the beginning of a mass disruption of, you know, what historically has been aggregators.So, you know, whether that's like the booking.coms or Expedia's, etcetera, because you can just have your agent look at, like, find all the best hotels in a certain area and have your agent go teach their websites and check out the best price. And you can even have your agent, you know, reach out to each hotel if you have enough time and kinda find the best price. So I do think this will cause disintermediation in places where previously kind of there were these aggregators that would essentially do the work for you since, like, you wouldn't really wanna go to 20 websites and get the prices of 20 hotels at 20 different websites, you know, so for the same time period. So it is exciting new use case. I think, like, travel is a good one, like I mentioned.Any big purchase, like, if you want to buy, like, a used boat or, like, a a house or something like that, you know, it's really great for kind of, like, searching the market and finding new opportunities. It's super good for social. Yeah. Great for used car sales to find an opportunity. If you're looking for a specific car, it'll, like, kinda search the market for you, let you know when new ones come online.It's good for social media both in, like, generating content as well as helping you, like, auto publish content. Obviously, like, market research, it's come out with some new tools like websites so you can, like, create research and then put it onto, a a landing page. I believe phone calls are coming soon. So, you know, it does feel like we're in a new age of, like, workflows that will come to the enterprise pretty quickly across, like, all enterprises. And, you know, whether we'll see, like, kind of the winners here, like, I think we spoke about it on last week's episode, but, you know, whether Gurak will remain a little bit or we'll see Closed come through with their own product or even companies like, you know, Google or Microsoft that, you know, kind of own the enterprise stack, you know, come through with a a better solution here that will end up end up being a winner.Rise Labs builds production grade AI for the world's most ambitious companies. So 40 in twenty years. Data scientists. AI architects. All working together on your problem.While other talk about AI Rice House is already implementing it.Sam Nadler: Why does the, like, the fact that it's working on its own virtual browser, virtual machine make it feel so different than, like, the previous iteration where, like, you know, Opus 5.5 would take over your browser and just do things on your behalf.Jordan Metzner: One thing is, like, having a coding assistant. So just, like, think of it as an employee who, like, sits next to you, and you tell them what to do and then they do it, and that is kind of, like, how Closed CoWork and what some of these other platforms work. Like, as long as you sit next to them, they'll do it. But, like, if you don't sit next to them anymore, they will stop working. They now have, like, pretty long run time, so you can get...And kinda get, like, a lot of juice out of it. But, like, it's kinda like I'm gonna go to lunch and I'll come back. I think what's different here is like because it has its own server, it can do these kind of cron job flows and workflows and extensive complex workflows and stay maintained and logged in and create state and things like that. So it's not that like you need your laptop open or your computer open in order for, you know, the Clod code to access your browser, to access LinkedIn, to make a LinkedIn connection. You just tell it to connect out to 30 people a day and it use its own browser to go connect and do all that.And that's like a huge change function from like kind of the necessity of you as a worker depending on these tools to like do the work for you. Right? One was like you had to guide it, the other is you can just let it go do its thing and it'll take care of it for you.Sam Nadler: Perfect. Alright. Give me... I think we were talking about how we've leveraged agents to, you know, update all of our websites. Is there is there something we can share there to our audience?Jordan Metzner: Yeah. I mean, not only is it able to like go into Webflow and make edits, but like make improvements and optimize it and, know, we are able... I built my own, you know, JavaScript tracker so that we can track traffic and make our own heat maps and understand user behavior on our site even better. I just read in our hero, and that just feels like there's constant ability to, like, make improvements both, like, on the design of your site, both... As well as from, like, an SEO perspective and then, like, on the ads perspective as well.Like I told you, kind of using these tools to, you know, generate AI based content and voiceovers and text to speech and using music and sound effects and all these other libraries in order for in order for it to make, you know, beautiful ads that honestly are highly converting. So, yeah, it's been really fun and exciting. And like I said, once the agent gets even more power and even more control here, I think it's it's like a step function improvement in in kind of where we're going from here.Sam Nadler: Let's jump into the news, Jordan. There's like really only one article I think we we hope to to cover today.Jordan Metzner: Yeah. Nvidia spy gate. It appears that Jensen was hanging out with a Chinese spy. And, you know, the story's crazy because, you know, you would think one, like, you would have so much more security around him. And just, like, way more protection.And then at the same time, just like, wow, China's crazy and, like, how important it is for them to understand what's happening at Nvidia and get this person in the room. And so kinda scary more than hilarious, but just shows, like, how deep, you know, the the war for AI and chips is and how important it's gonna be moving forward. So anyway, Sam, this was a great episode. Thanks so much. It was awesome.I think go check out the bots. Thanks for enjoying Bot This Week.
