Loading summary
Transcript75 lines
- [00:00]
A
Rowan Chung is the genius behind the Rundown. Now, the Rundown AI is a media company for AI that's generating millions of dollars a year. Now, I thought that the rundown had like 50, 100, 150 employees. No, it's a lot less than that. He just uses AI agents to actually scale the business. In this episode, he unveils it all, and I think it's for the first time ever. How he created an AI avatar that gets me millions of views. How he writes tweets with AI that gets millions of views. How he writes a newsletter with AI that gets millions of views. How he created an EA AI agent that answers leads within seconds. How he's able to learn anything from a YouTube vid with AI, how he onboards employees with GPTs, how he built a partnership AI agent that works directly with his team. We give away all the prompts, all the tools, all the workflows, all in this episode. I hope it gets your creative juices flowing. Stick to the end. Give us a like comment and I'll see you at the end of the episode. Long time coming. Rowan Chung on the pod from the Rundown. My go to guy when I need something explained in AI. Thanks for coming on, Rowan. By the end of this episode, what are people going to learn?
- [01:29]
B
Yeah, thanks for having me. I love the show. I love your content, so it's great to be here. By the end of the show, I'm going to share seven ways we use AI internally at the rundown. They're going to be business, operation and content creation focused. We use AI many ways, but these are solid ones that people can walk away with. We're going to share some free templates that people can steal, and then I think we're going to riff on just some general AI overview for the future.
- [01:58]
A
So when you say like operations and systems, what I think you're probably talking about is how do you. How do you take a team that's maybe 1, 2, 3, 4 people and make it seem like it's 20, 30, 40, 50 people? Is that right?
- [02:12]
B
That's exactly right. So we have a team about 15. I think we operate as like a 50 person team in terms of speed, but in some cases we might actually operate faster because the headache of having 50 people is really burdensome. But yeah, we just use AI in so many different ways that we operate extremely fast and we can get a lot of work done.
- [02:35]
A
Okay, cool. Well, I'm here to learn just as much as everyone else. I think a lot of people who listen to that They've seen other tutorials or videos and they're like, yeah, right. There's no way that AI can make a team of five turn into 50, but if there's anyone who could prove it, it's you. So let's get right into it.
- [02:53]
B
Yeah, let's get into it, man. So we can start with this first one. I think this is probably the most popular one. It went pretty viral on Twitter, even though, I mean, this is Instagram. But basically the problem we were solving was I have no time as a founder of, you know, the rundown. We're growing our newsletter, we're growing our edtech side and I also do content on the side as well. And I'm a writer first. Like, I'm not great in front of the camera. I'm getting better, trying to get better. But we realized, like, we're creating all this content in the newsletter and we have all these stories and it's so in demand right now with AI content. And we needed a way to push that writing into like some short form video. At the time, I didn't have a setup, I had nothing. So it was like, okay, I don't have any time to do this. I'm not going to do my hair and get in front of the camera and record this five minute video, do six different takes. It just wasn't going to happen. So we're like, okay, what if we run an experiment and we clone my face, we clone my voice and we just have an editor who takes my original writing and pushes that into cloud script so we can, you know, form. Formulate that better for like short form and then run that through the avatar. Then obviously it takes a lot of manual editing to make the pieces right. We got to do a lot of storytelling. There's a lot of like B roll that we have to add. But we said, hey, can we do this? Like, this is going to be experiment. We're very open about it. So on Twitter we're like, hey, we're running this experiment. We're not trying to fool anyone. Like, this is an AI avatar. But if anyone were to try this, like, we're gonna try this and if it works, like, it's a cool use case because I don't have any time to create content on short form anyways. So, yeah, we started doing it and it's been under a year and we've grown to over 150,000 followers. So I'd say it's been a success. Yeah, you can go through any of these. Some of the recent Ones we've been experimenting with myself narrating it as well, just because I now have a studio, as you can see. But I didn't have a studio before, so we just used that avatar and it's way better than me anyways. So you kind of go through these. We had a few, like really viral ones.
- [05:03]
A
Sam Altman, the co founder of OpenAI just said that it is the era of the idea guy and he is not wrong. I think that right now is an incredible time to be building a startup. And if you listen to this podcast, chances are you think so too. Now I think that you can look at trends to basically figure out what are the startup ideas you should be building. So that's exactly why I built ideabowser.com and every single day you're going to get a free startup idea in your inbox and it's all backed by high quality data. Trends, how we do it, people always ask. We use AI agents to go and search what are people looking for and what are they screaming for in terms of products that you should be building. And then we hand it on a silver platter for you to go check out. We do have a few paid plans that, you know, take it to the next level, give you more ideas, give you more AI agents and more. Almost like a chatgpt for ideas with it. But you can start for free idea browser.com. and if you're listening to this, I highly recommend it. So 120,000 likes. I actually saw this one. Yeah, this is talking about Anguilla and the AI domains being like the number one revenue source for them. Is that what this is?
- [06:20]
B
Yeah. And you can see like, um, if we reload this, the avatar is only really shown for the first like 3, 4 seconds because the avatar portion of it isn't that great. But if you just show like 3, 4 seconds, it looks more real. The audio, on the other hand, like audio is there, like voice. AI is really advanced. I'd say video is like lagging behind a little bit. So the voice is also AI generated. But I'd say the voice is sometimes it's better than me even. Like, I mispronounce things all the time. But yeah, that's basically the way we do it. And I have an incredible editor who, you know, does all the storytelling. He pushes it all together. He's taking these stories from, like, things I've written in the past. And yeah, it's like a cool workflow that works really well for us. I can't say it's going to Work well for everyone. Like, it does take a lot of work and you do need like a really talented editor to be able to do this. But for like those edge cases where you have no time as a founder and you really need to create content, short form content, and you're already writing, it's just kind of like a no brainer. Right. So yeah, that's the, that's the first one.
- [07:34]
A
Yeah, I think, I think the mistake a lot of people make here. Well, there's a few mistakes. One is they, they don't pick stories that are going to go viral within short form. Right. So they, they, they make the mistake of just picking, I don't, you know, boring, boring content and wrapping it. You know, you can have the most boring content, but if it, even if it's wrapped in the most beautiful short form, as you know, it's not going to go anywhere.
- [08:00]
B
Yeah, exactly. And the other thing is like a lot of people think this work, they see this workflow and they think it's like one click to generate. We do a lot of work. Like my editor even says sometimes it would be faster if I just came in and I just like quickly narrated it, but I don't even have that like 10 minutes of time. And like that break in the workflow every single day takes a lot of like mental capacity. So it really works for us. But I think a lot of people get that wrong where they're like, hey, we're going to one click generate this and it's going to be awesome. And they don't spend time training the avatar in. Hey, Jin. I'd even mentioned the tools we use. So we use Heygen for the avatar and we use 11 labs for the voice. Heygen has like 11 labs integration, but you actually want to do it separately because if you do it on 11 labs, you get higher quality and you just stitch it together. So those are the main two AI tools we use for this workflow. And yeah, like I said, it takes a lot of editing between, but it is really useful for people who do get it right and get those stories right, like you said.
- [09:03]
A
Yeah, and you can also, I mean, you can use tools like, you know, Gumloop and Lindy and stuff like that to actually pull in stories that are getting traction on places like Reddit and then so it's easier to kind of pick the stories. I know you're actually creating the stories, but most people listening to this probably aren't writing how many thousands of words you write a day.
- [09:27]
B
Yeah, that's a, that's a good workflow. Um, and the other thing is like just being active on X. Like things break, break out first and I'm like, me and you are both probably there like 24 7. So we are really on the pulse. But yeah, like if you're not on the pulse of things having something like a gum loop integration where it's like even pinging you on Slack for the top stories that are breaking out right now, that would be really useful. Yeah.
- [09:54]
A
Cool. Before we move on to the next thing, you want to show why shouldn't someone. People listening to this are starting companies, they're starting businesses, they're starting cash flowing businesses. Why shouldn't they do this?
- [10:09]
B
Well, I think I actually got asked this by a public company. He asked me how to do this and I went to his office and I explained the whole thing and he's really interested. And in his case I actually think it was somewhat useful because he's actually from Switzerland and in Switzerland there's four different languages and he has to record these videos in four different languages every single day. And it's like it takes up hours of his day. So for him it kind of is useful. But for like the regular everyday CEO, there is some element to it where there is AI hate. Right. If using an AI avatar for creating your videos, people are gonna think, oh, it's like not as authentic. For me it really worked because we were very open about it and I'm kind of like the AI guy. So people accept it. They're like, oh, this is really cool. He's actually practicing what he's preaching. But for a publicly traded company CEO probably shouldn't do this, should probably be as authentic as possible and strive for the human element and really like push for like that human experience. Because like as we get into the age of AI generated content and there's just like seas of a generated content, you want to push for like that human layer because that's really your moat. But for us again like it works for our specific case and there's some cases that it will work. But yeah, like, like you said, there's definitely some areas we don't want to use this.
- [11:42]
A
I mean, I think what's so smart, and I just took some notes on this, I think so smart about your approach is the video avatar piece is like 2% of the video. You know, it's. And you're just using B roll. So I think even if you are a public company CEO, like if you're like me or you, by the way, people think it's funny because we're on video a lot, but people don't realize we actually don't like video. Like, we're. We're actually like, I'm afraid of the camera. You know, I've tried to do short form and I tried to like deadpan into the camera and I'm like, not good at it. And I don't. I just honestly don't. I don't feel good after. That's the best way to describe it. I don't feel good after. I'm not good with the camera. So I think there's a lot of people like us who know that there's so much value in short form, who know that if we actually just looked in the camera and spoke these stories, that we know we would actually crush it. And that, you know, if you climb Cringe Mountain, you could get to 151,000 followers and just like pick up millions of views. That can absolutely change the trajectory of your life. That being said, we don't want to do it. So I think that for anyone who's listening to this, this is an interesting approach, at least to try as an experiment for 30 days and then see how you feel about it. Yeah.
- [13:05]
B
And another case on like, the pros side of things is faceless channels go viral all the time and they generate millions of dollars for people, right? Or even you see these animated characters on YouTube that. It's like this trend now as well. And you can build a following in a, in a community through animated content. Why can't you do it through like an avatar? And is this going to be more popular as time goes on? Right now you got to stitch things together, right? It will get a lot easier and that means more people will do it. So we'll see what the world really thinks about it as things get more popular. But for now, it's definitely a huge arbitrage opportunity. And we grew the account to 150,000 and it's working for us.
- [13:55]
A
Well, thanks for sharing that sauce. What else you got?
- [13:58]
B
Yeah, the next one. I mean, this is a pretty simple one. I'm sure you do this all the time as a writer, but it's just voice dictated writing and you push it into a cloud, like workflow, so there's no magic secret other than this prompt. So I mean, people can. We can like drop this in the description, but it's basically just a prompt to train on your writing style, right? So I grabbed all of my top performing tweets on X ever, right? I have like 10 to 20 of them. I drop this prompt into Claude and you basically just push all of your best writing and it's going to be trained on your tweets. And then I'm. This is a pretty popular workflow, but just like, I like going on walks and talking on my phone. I got to learn my best ideas while I'm walking. So I just, like, walk and I just have whisper flow open and I just like, jam on different things and then I have a bunch of different notes. And then when I come home, I just, like, push into this chat that's trained on my writing. And then you just get really, really good tweets. So, like, for this, this example, they're pretty optimized tweets. You still have to go in and edit it a little bit. Like you want to get this. So once I get this, I'm going to push it into, like, Twitter, for example, and I'm going to add a couple lines that are more like, give more insight. Right, the insight part. AI sucks at giving human insight and experience. Comparing certain things to other models, for example, and just tweaking up little things that just don't seem right. Like, Claude likes to exaggerate a lot. Like, I say words like game changer. But this is like a really, really good starting point where it takes you from, like, walking to literally tweet is 90% ready. And I, what I do is every single Sunday, I just go on this really long walk and I just jam on my phone and I have a bunch of different voice notes and I just like push it into here and I'll have all my tweets ready by the end of, like, Sunday and I'll like cue them all up and then. Yeah, it's like a really good workflow that works for me when you want to push out a lot of content. But yeah, do you use something similar?
- [16:18]
A
Well, I have a question on your walk piece, so do you just go on a walk and ideas hit you or do you have like. I know I'm speaking about myself. I have an Apple Notes called Ideas, and in my Apple Notes, I have these kind of half bake ideas that then I'll use whisper flow to actually dictate it in. But I'm just curious, do you start from nothing or do you start from something?
- [16:47]
B
I usually start from nothing. The ideas really come to me when I'm walking. I might have some broader things that I'm just thinking about. And I don't know, I think I work so much that I don't have thoughts until I start walking, my mind is almost numb because I'm working so much. So once I start walking and like, I'm not looking at slack, these ideas just come to me and I just, just start jamming from there. And yeah, it's a really good workflow for me to, to just have tweets, like 90% ready when I come home and start queuing them up. Obviously I'll add like an image or some kind of video because video content does really well on Twitter. Uh, but yeah, this is like a pretty simple one that I think will be useful to a lot of people who are trying to create like, LinkedIn content or, or Twitter content.
- [17:37]
A
On my last question on this workflow is you mentioned this is really good at getting, you know, 0 to 90%, but the last 10% is everything. So what are, you know, what are the prompts? You, you, you kind of mentioned a little bit of this, but what are the prompts that help you get it from 90 to 100, or is it just you?
- [17:58]
B
A lot of it is me, but you can go back and forth a little bit. On the next example I'm going to share. It's basically the same thing, but I use it for my newsletter. I'll share how we go back and forth and use it as an editor. But yeah, a lot of it is you take the tweet, you add in your insights, you refine it, and then you ask again. You're like, hey, how is this? What do you think? Make it better. And then it's going to give you some suggestions and again, you don't use them all. But the one suggestion that it might suggest that you didn't think about could be like, the reason why that post goes viral. Right. So it's just like, you have to develop this intuition of knowing when it's a good suggestion versus when it's not a good suggestion. It just comes with reps. But yeah, Claude is amazing as an editor.
- [18:51]
A
And do you ever have. Is this Claude Max?
- [18:55]
B
That's actually a long story. So I actually can't get Claude Max because my account was upgraded by Anthropic and it was like a whole thing. We're trying to figure it out. But yeah, this is just regular Pro.
- [19:08]
A
Account and you never have the issue of, I know we're getting to the nitty gritty, but I actually think people want to know the nitty gritty. You never get to the issue of, like, the chat is too long, therefore you can.
- [19:23]
B
Yeah, that's. I mean, this is a huge issue that Everyone always has problems with. I'll get to like the. So this is the next example I was going to share. It's basically this exact same prompt, but we use it. This is a different like, use case. We use it as an editor. And this is exactly what you're saying. So you can see like how long this chat is. Right? Like, this is insane. And it's really like, when we write our newsletter, we write it all by hand, but we like to use Claude as like an editor in chief. Right. And it's, you're going back and forth with it, you're asking for suggestions, you're saying, oh, how is this? What do you think? Like, can we make this better? It's going to give you some ways. And you say, oh, can we make this a little bit shorter? Can we make this two words shorter? Because that's a fit in like our email format. Right. And Claude's extremely good at this. But to your point, when the chat gets too long, Claude's just going to say, hey, this chat's too long. You can't generate anymore. There's a pretty simple hack. You literally just go to. You literally just go to another chat somewhere not too far up. You want to make sure there's enough context still because the chat, like Claude, actually gets better as it goes. But you want to go to here, for example, you see, I have two edits and you literally just go in here and you can exit that out and you can start from there.
- [20:47]
A
Wow, I didn't realize that.
- [20:49]
B
Yeah, you can literally just start again, just a tiny bit up and it just keeps going. So especially if you're using it as an editor. For example, we go to my newsletter here and we grab like a main, main pair. Obviously we already published this, but if we were editing this still, we would write this all by hand and we say literally, like, this is literally the prompt. Make it better. It's super simple. But you hit save and it's just going to start from here. Like all that bottom part gets erased and you're going to be able to continue the chat.
- [21:27]
A
So do you have to look for like a one on two situation? Is that what you have to look for, like, or could it be.
- [21:34]
B
No, no, it can be anywhere. I, I'm just showing that example because it's kind of, I guess it's kind of proof that I, I genuinely do this. Like now it says three, right?
- [21:42]
A
Right.
- [21:43]
B
But if you go further up, you'll see some where it's like 10 or 12, because I just like stay on this chat, I've trained it, it works well. You don't want to retrain every time. It's kind of annoying.
- [21:53]
A
Yeah.
- [21:54]
B
But yeah, this is basically the workflow. Like it's going to give you a refined version and you're not going to copy and paste this. That's the worst thing you can do. But you're going to read this and you're going to have your article open on the other page, you're going to compare it and you're saying, oh, that's a better way of actually explaining that bullet point. Let's grab that and refine this word. And it actually takes more time to edit with AI as a co editor. But your content gets so good because it's trained on your best writing and you're just going to find these small details that you wouldn't have otherwise seen. Like even if you hire someone, it's like it takes so much time for them to go through it all. And this you're really in control. So you have like an editor in chief that you're in control of. And yeah, it's just been a game changer for us. Writing content newsletter. But also just like writing content on LinkedIn, Twitter. It's a really good way of doing it.
- [22:53]
A
Cool.
- [22:54]
B
Yeah. So the next one, this is a more fun one but we use Lindy. So I know you had Flo on the pod. He's great. I love Flo. But this is just a meeting schedule agent. So Lindy actually provides like a template for this but I've modified it a little bit because I wanted like my workspace email. So I actually created an agent and I called it Ava. So Lindy has this like they give this own template and they use like their email domain but I basically switched it up and I switched a lot of the AI prompts to basically use my workspace domain. So basically what this does is whenever I'm in like a Gmail chat here, I can actually show you an example. Whenever I'm in a Gmail chat, I can literally just CC in Ava. So avandown AI and then Ava will come in and like literally look at my Google Calendar and find available times and we set that parameter of like where I actually want available time. Because like you don't want your full 40 hour work week available for meetings. Right? You want to have some deep work blocks. So we scheduled exact times of like where I actually want to take calls and it will actually look through your calendar and request those exact times and say, hey, did any of these Windows work for you. The other person can respond, say that time works, and then it will go ahead and book and it'll actually book in your calendar. And then this whole section here is basically like rescheduling. So it's. Let's say this person came back in the thread and say, oh, actually this didn't work. This. This time didn't work. Let's reschedule for, I don't know, like the last time here. Then Ava will literally go in and cancel the event and then reschedule another meeting. And it's just like a really nice way to, I don't know, not send a calendly link.
- [24:59]
A
Yeah.
- [25:00]
B
So yeah, it's a cool little use case and you can actually take it a step further. I don't have it in this workflow. I have it in another one. But you can take it a step further and actually enable this across your whole team. So your entire team can loop in that one email. So avaterundown AI and Ava will be attached to their calendar and you can just literally, Ava is the meeting scheduler. It's the go to meeting scheduler. And you don't need calendly. And I don't know, there's this weird stigma around calendly. I don't know how many people who watch your podcast know about this. I don't know why I don't agree with it. But, like, if you send a calendly link, it's kind of known. I don't know, it's kind of disrespectful almost. So having an agent that books the meetings for you is just like a better way to work around that. And it's useful for your whole team. So, yeah, it's a cool little agent we built.
- [25:50]
A
Go back to your email. So let me explain why I think that this is such a game changer. So a lot of people listening to this sell a service or products, you know, but a lot of people listening to sell services. When you're in the services game, time is of the essence. So if you know, who is this person? Nicole. So if Nicole is like a lead, basically you want to get this time booked as quick as possible. When I'm looking at. Exactly. When I'm looking, you know where I'm going with this. When I'm looking at the timestamps, I'm looking, you know, basically you're responding instantaneously with the Lindy AI agent. It's a minute. It's a minute. Yeah, one minute. And even if you go down, scroll down I even think, yeah. 6:42. Nicole goes, hi, Ava. Tuesday at 7:30 works for me. 6:42 responds. Right. Ava responds, sounds great. I just sent the invite so that the, the gap, like if you would have 6:42pm your EA or human AA might be off, you know.
- [27:04]
B
Yeah, it works 24. 7.
- [27:06]
A
It works for you 24 7. So I think that piece of it is so valuable.
- [27:11]
B
It definitely is.
- [27:15]
A
I mean, and for you too, right? Like even for like, you know, you guys sell sponsorships for example, and maybe like Microsoft, you know, CMO of Microsoft comes in. Right. And you want that meeting, you know, also might be someone interesting. So you want to make sure that you know, these busy people, that time gets booked.
- [27:35]
B
Yeah, yeah. You can actually use this for partnerships. So we tried this for some time. We turned it off recently, but we tried this for some time where we have this workflow where we have a form. Well, we have, we obviously have big distribution. So a lot of our leads come internally inbound. So a lot of these people come inbound. They answer a form. We get all these questions and we use those questions. We push into Lindy and it basically validates them. And then based on that, if they're validated, Lindy goes ahead and like sends a sponsorship email right away. And then they can literally go back and forth and if they want to book a call, then they book a call. With my human salespeople. It worked for a while. But yeah, we're getting like the use cases where you, when you do get to like Microsoft and like Google, I'm just a little, a little weary of using agents right now because there are like occasional times where like it could go wrong. And just that like 1% knowing that it could go wrong 1% just worries me for those like really high tier clients. But when you have so much inbound, it's, it's extremely useful. But for this meeting use, meeting schedule, use case, it's, it's perfect. So we can like drop this template as well. I'm happy to share it in the YouTube comments or something.
- [28:54]
A
Yeah, let's do it.
- [28:56]
B
But yeah, anyone can just steal it. Like literally all you have to do is connect your own email.
- [29:02]
A
Yep.
- [29:03]
B
So you probably don't want to use AVA at the rundown AI. So you're going to want to change that. And then here's where you basically add all the emails of your team. So this is my, so this is the one with only mine right now. But you change your email and then you want to connect your Google Calendar and then in here, the only thing you really need to change for these two send replies is at the end it says Best Ava. So if you change your. If you change the name of it, you don't have to call it Ava, obviously, but you can change the name of it. And yeah, you can call it whatever you want. And also here, sorry, one more. You want to make sure you change when your meeting hours are right. You don't want to have the whole day, otherwise, you know, people are going to schedule random times. So you for me, only have three hour blocks, four days a week. So that's like my meeting times. So, yeah, those are really, that's really it, what you want to change. But yeah, we're happy to drop that in the chat.
- [30:01]
A
Appreciate that.
- [30:02]
B
All right, so the next one, this one's a kind of research use case and it's with Perplexity Comet. So did you get access to Perplexity Comment?
- [30:14]
A
I did, I did, yeah.
- [30:16]
B
So it's pretty cool. My stickiest use case is using it as an AI tutor for like every video, like long. For example, this video itself, right. If I don't want to sit through the whole thing and I just want to see Rowan's seven use cases, tell me exactly how to do it. Perplexity can get you that in like two seconds. Um, so it's incredible use case for just summarizing these long videos. And it's really sticky. Like, I really like this use case. But it's also useful for just going deeper on topics you don't understand. Right? So for example, I use ChatGPT a lot or ChatGPT voice when I'm reading. So I'll just have it open next to me. And whenever there's a topic I don't really understand, I'll just ask it, oh, hey, what does this mean? And I'll go deeper with it before I move on to the next page. And it really helps, like retention. But Perplexity Common is basically that for every YouTube video or podcast that's on YouTube, so you have this widget that's on the side of your tab, your comment tab, and you just go into like these mini rabbit hole sessions and you can understand things way deeper than if you were just watching the video. So every video in essence becomes like an AI tutor. So yeah, I think it's pretty, pretty incredible. I learned so much better like this. I never did well with school, so YouTube videos are always my jam. And this is like a huge game changer for me.
- [31:50]
A
I haven't tried. I Honestly haven't. I've been playing with Assistant like all day, every day, like Perplexity Assistant, but I've never tried it with YouTube videos, so I'm definitely going to try it. I'm like, you also like, I learn from YouTube and like not surprisingly. And I also, I've got like my Mount Rushmore of people I really look up to and like, you know, Dharmesh is awesome in this example, right. I would love to learn about AI from, from Dharmesh. Right? And I could do that with Assistant or, or even like imagine someone listening to this, right? Like with the Lindy AI example it's like, okay, this is, you know, I have a question about what Rowan said about Lindy templates and then you can just post it there and just have a full on conversation with basically AI Rowan using Perplexity Comment.
- [32:50]
B
Exactly. And the cool thing is the comment Assistant is grounded at the entire context of the video. You can really ask it anything about the entire video and it's going to be smarter than a regular Perplexity would because it actually has that context. So exactly like you said me and you might not have time to respond to every single comment like of this YouTube video, but you can ask Comment and Comment will probably have the answer and might actually answer better than we would. So yeah, it's incredibly useful. I've been using it as a tool so hopefully it's useful to other people as well.
- [33:26]
A
I mean, doesn't DIA have their version of an Assistant as well via the browser?
- [33:32]
B
Yeah, they have. I have not used it as much as I should have, to be honest. Comet came out like right after and I'm like, okay, let's get on this. I heard DIA also does this, but I don't know, I like Perplexity and it's just really sticky. Like I don't really want to switch at this point. It works, right?
- [33:51]
A
Yeah. Let us know in the comment section if people have used DIA and that works or if you prefer comment.
- [33:59]
B
Yeah, I'd love to know as well.
- [34:00]
A
Yeah, I'm just curious, was there one more you wanted to share or was that it?
- [34:05]
B
Yeah, I got two more. Two quick ones. This one is a GPT for new hire onboarding. So problem operating new hires is extremely, extremely tedious. When you're scaling your business, right, everything's growing really fast. But you need new people to take over parts of your business so you can scale higher. But like teaching them the ins and outs of your company, teaching them the processes, teaching them how you use AI, teaching Them the tools that you use. There's so many questions and it just slows everything down. Right. So we built this GPT that's trained on SOPs, literally loom transcripts, meeting notes, our policies on AI usage, HR docs, any like admin related question. And it also knows everyone on the team and their, their role. So basically the motto then becomes don't ask questions. You haven't asked GPT first because the GPT literally knows more about the company than I know. I don't remember all of these policies even. So you can literally ask it anything about our company. You can ask it if we can take a vacation today and it'll like answer it and then if it doesn't know, it will point you to the right person on the team to ask. So then you go and ask your manager. But yeah, it's a little, little hack we use to basically scale faster, easier, save time for the managers. And yeah, it's, it's pretty useful. We can go into the back end here. So for anyone watching this, if you just want to grab this prompt, it's a pretty simple prompt, but really it's just, you're making GPT into a custom onboarding assistant and then you just want to drop as much knowledge as, as you can and you can always update this, which is the cool thing. So as you get more team members as you, you know, create more HR policies, that sort of stuff, you just drop it in and the GPT just gets updated. And it's really useful.
- [36:23]
A
I need to do this. I need to do this. It's useful. It's also one of those things where some people are afraid to ask these questions. So they don't.
- [36:33]
B
Yes.
- [36:33]
A
Yeah. So I think that for a lot of employees this will just be so helpful for them that they can just any time of day, night, they don't have to, you know, look, you know when you're asking a question, especially if you're first starting out, you don't want to be, you want to make a good impression. So I can totally understand where people are coming from there. Yeah.
- [37:02]
B
And really like this thing answers questions immediately. You don't have to wait for your manager to respond. And like someone like me, I have so many other things that I'm trying to remember that like I don't remember all of our policies. So if someone asks me a question, I'm gonna have to go in and look in our drive and like look through the points and be like, oh no, sorry, you can, you can't take this day off. But you can take next week off.
- [37:24]
A
Yep.
- [37:25]
B
But like this assistant can answer that immediately. So it's a pretty useful thing as you're scaling your business. And then the last use case I want to share, this is a little bit more a niche use case but it actually probably be useful for you as well. But for anyone that has a ton of replies to like their business email inbox. So for example we have two main inboxes that we use this for our support inbox and then our newsletter email inboxes. So so obviously we get a ton of people who just reply to our newsletters and we get a ton of out of office email responses, a ton of spam. But like there will be really good feedback in between some of that junk and that's gold. Right. So we, we need that and we need to respond to those, that feedback cause those are like our, our winning readers or winning customers. Right. So this agent does three things. One, it filters out the real genuine customer support in the inbox so it like filters out all the spam. And then two, it pings the right people on my team in Slack if there's anything important. So for example, if there's something sponsor related, it will go ping our sponsor team. If it's something education related, it will go ping our education team. If it's something like deliverability or I don't know, someone tried to unsubscribe but they couldn't, then it will ping me. So it's automatically pinging people on the team and then the last thing it does, it actually auto draft replies in the email inbox. So when we go in to respond to some of these people, the responses are like nearly there. It's the same thing with the voice dictation workflow where it will be like 90% there. We still have to sometimes add context, but it's quite good at saving a lot of time. And this is the only way we found that we don't miss any of those like really important emails in between all the spam and the out of office emails.
- [39:32]
A
Yeah, it's. I have this issue I actually have on my YouTube. I have my email and I'll be honest, I don't even really look at it. So I'm getting like tons of emails.
- [39:44]
B
For a long time we didn't look at it either because it's like you're getting hundreds if not thousands a day. It's impossible. Right. But there is that really good feedback in the middle. This just handles all of it. It just runs one time per day as well. So it's not going to be using up all your Zapier credits. It just runs once per day and then it's just going to ping you in Slack if there's anything important. So, yeah, you know, if there's anything important. And we've actually landed two sponsors through this. Like, it's really weird that they don't go through our actual sponsor form, but these are legit sponsors from like big companies that went through a support inbox. Right. There's, there's really important messages that come in there. So I'm happy to share this one as well. It's pretty simple to set up. It's using Zapier agents, so it's like their agent feature and we use actually their, their copilot. So we just literally just drop this prompt in their copilot and it actually builds the zap out for you. Like, you actually don't have to do anything if you use this prompt. So we can share this prompt in the chat if anyone wants to use it. You just got to drop this in the copilot. It'll automatically generate all the connections and then you just have to like connect it to your Slack or email and all that stuff. But, but it's ready to. It will be set up in like two minutes if you have this prompt.
- [41:07]
A
Beautiful. This is helpful. Yes. All of the stuff from today will be in the, in the show notes for people to access. And I appreciate you sharing that, man. A lot of people actually come on here and they're like, no, I'm not sharing that. But Rowan, you're the type of guy, you know, you're a sauce factory. So that's why I have you on here. And thank you for, for being so generous with, with, with your time and, and, and your. Yeah, the whole everything.
- [41:39]
B
Of course, man. And again, there's so many other ways we use AI. I tried to stick to the content creation and kind of unique business operation ones, but yeah, we use it across like coding. We have marketing workflows, we have stuff we do for education as well. So yeah, happy to come on and jam any time and jam on some other stuff, but hopefully this is a good starting point.
- [42:03]
A
Yeah. If you want Rowan to come back on, we got a scream to come back on. So let us know in the comment section. I'm going to include places where you can follow Rowan AI Rohan Non AI Rowan. In the Show Notes X newsletter, Instagram, I'll include that. Is there anything else, Rowan, you want to share before we head out?
- [42:27]
B
No, that sounds good. I'm mostly active on X. And yeah, if you don't subscribe to the rundown and you want to keep up with everything in AI, it's the easiest place, so we'll put that in the notes as well. But yeah, it's been great coming on and I'm a huge fan of the show, so, yeah, it's been awesome.
- [42:46]
A
I want you to come back again, man. I want you to come back again. I want the comment section to be roaring so that you'll know that it's not just me. Right. It's the people. The people want you on. So I have a feeling that they're going to like this episode. And I'll see you next time.
- [43:04]
B
Cool. Yeah. Appreciate it.
- [43:06]
A
Later.