← Marketing Against The Grain
Loading summary
Transcript54 lines
- [00:00]
A
Welcome to today's show. This is a wild one. We're sitting down with Riley Brown, who is one of my favorite AI thought leaders and experts and he's taken us behind the scenes, how he builds agents skills, how he works with them across his company, how he even went through his entire six year iMessage backlog and was brought to tears. This is a really powerful, really interesting show. Let's get right into it on this show. We're here with Riley Brown who is content creator, AI educator, somebody I follow very closely on X.com and the Internet for a lot of AI insights and advice. And Riley, you're one of the people I've been really excited to have on the show because I just think that you have very clever ways that you build and use AI and I just wanted to bring that to our audience because I find the most interesting thing about AI is to learn how other people are using it and like actually better understand what's possible. So I'm super excited to have you on the show today. And maybe just so start with how the heck are you using AI every day? Because you are using it to the 99th percentile I think compared to most people or at least most non technical folks that I talk to.
- [01:09]
B
First of all, thanks for having me on the show. I'm really excited. I guess the way I use AI is very dynamic. I'm constantly asking AI to do something and then I'm having the AI take whatever I do that's useful and I turn it into a skill that I can repeatedly use. And I think that is step one. And in fact I'll just show my screen real quick.
- [01:27]
A
Well, please basically taking everybody on a tour inside your Codex and how things are actually getting built.
- [01:33]
B
Let's go ahead and when you open up Codex, right? And actually let's talk about what is Codex. Codex is OpenAI's version of Claude code or their Claude desktop app. And all of these Frontier Labs or the, the big companies like Anthropic Xai, OpenAI, they're trying to create these apps called super apps. And this is what I've been talking about for a while. This is a place where you can interact with AI agents or you can use AI agents. And the number one thing that I think you need to understand is like this agent can basically do anything on your computer. And so for me I have a ton of these skills and so I've created all of these different skills which are just repeatable workflows that I can just use at any given time. My favorite skill that I use and that I'm constantly using as a YouTuber is this YouTube researcher skill. And this YouTube researcher skill uses an API, it's called Supadata, to really quickly pull YouTube transcripts. And it can also search YouTube like I could have it go check all of your episodes, rank them by popularity, and then it will only analyze the top 10 videos. And so I could say analyze my last 10 videos and give me advice. And this is a very simple prompt that I can use at any given time. This whole YouTube researcher skill, it is an instructions file that it can use at any time. And I think if you're a marketer, anytime you use AI for something useful, and whether you're retrieving data or whether you're creating a certain type of asset, you can just ask Claude Code or Codex in this case to turn it into a skill so that you can use it over and over again. So I think that's kind of a good step one.
- [03:14]
A
Your point is basically, hey, if you take the time to build a skill, then the prompt is actually pretty basic and you can leverage that skill, which has a lot more context than you would put in a one off prompt. I think this is to be one of the things people actually struggle with when it comes to AI Riley, is that it's not using AI, it's understanding how they work. I think a lot of people, the step they're missing is like, oh, I have a YouTube channel. I need to analyze it on a regular basis. Because of that, I need a repeatable way to get all of the data I need to analyze and to set instructions for how I want that data analysis to actually go. And so I think one of the most important things people need to do is actually document what are the things they do in an average week, every week, because they need a foundation of these skills that's really going to transform how they use doesn't be Codex cowork In any AI model, right?
- [04:07]
B
Yes, 100%. And to kind of dig a little bit deeper in that, one thing you need to understand here is like notice here it's kind of like taking a while. And yes, this is kind of the future of using agents is multitasking. And this is one thing because the longer agents take, the more it's going to feel like delegating to a coworker. This is a short task. I've often found times where I'm working on four different chat sessions with agents at the same time. Each one is taking around 10 minutes. So really productive People are kind of forced to parallelize a little bit and do different tasks at the same time. That's the first thing I want to say.
- [04:41]
A
That's a win for all of us that are kind of ADHD out in the world.
- [04:45]
B
Oh, adhd.
- [04:46]
A
It's a little hard though, if you're somebody who just likes to go deep on one thing for a long period of time. I do think there is some working stylistic things that are different in this new AI world.
- [04:59]
B
100%.
- [05:00]
A
Thanks for watching today's show. If you want the best AI resources we've got for you this week, you want to scan that QR code or click the link in the description below. And we've got an awesome set of stuff for you.
- [05:11]
B
One thing I will add. Let's say you're a marketer and you were saying like, what do you do on a day to day basis? And usually that's actually too high level. What I've been doing is I've just been taking every, every granular task that I do and really trying to write it down. So I create a YouTube video, right? And this YouTube video should be good, right? And when. But. And what I mean by good is well researched and with like a proven format. And so One thing that YouTube researcher skill does really well is it can go search similar channels to me. And then what it'll do is it'll find other people's formats, right? And a format is just kind of like their hook and kind of their setup to the rest of the video. And that's what I call a format. And it'll look for formats and it can basically port that format over to an outline. So if this is the result, step one is like coming up with an idea. And I have a skill in Codex called brainstorm ideas. The next step is like creating what I call a hook outline. My videos aren't fully scripted, but the intro is highly scripted.
- [06:20]
A
I've learned this the hard way, man. The hook scripts are everything.
- [06:24]
B
Everything. And so I have a skill. It's called hook outline, right? This is one of the steps that I do. And I will say, like, Fable just came out. I need a hook outline for a Fable News video. Please look at my previous agent native episode and use a similar format. I need a hook outline. Oh, and then this is kind of the next phase after Skills is like put this in notion, long form, in progress and so codecs can fully control notion. And the coolest part about Codex, and this is what separates it from Cloud Cowork. It'll Create a notion doc and it'll give me a link for it and I can open it directly inside codecs. It has a full in app browser, which I think is the future of using agents. It's just like you put a browser inside your agent platform, you don't use some external browser. It is such a better experience. So that's kind of my one thing that I recommend learning is learning the in app browser inside codecs. And I think Claude desktop is going to add the same thing and it's going to be huge.
- [07:34]
A
I think what most people do is, is they work in regular ChatGPT or Claude and they just ask questions and don't build anything repeatable and they don't integrate it with any other steps in their workflow. And the biggest thing that you're showing is like, hey, I have to make these videos all the time. And because I have to make these videos all the time, I need a set of repeatable skills. And what I love is that you just showed the very basic thought process of like, here are the steps I need to follow. Like I need a hook outline. And if I always know I need a hook out line, then I need a skill that's going to go and do that hook outline. Right?
- [08:08]
B
Yes.
- [08:09]
A
This is what most people, I think don't do is they don't document the kind of core first principle, basic foundations of their work and then say, all right, I need to actually invest in building the skills and infrastructure to do these things I have to do all the time and do them really, really well.
- [08:25]
B
And one of the reasons why I think people are kind of struggling with using AI in marketing is they're just not good at some specific skill in marketing. And that's one thing that's incredibly important, is you have to get good. Like everything is a refinement. Right. When it creates a hook outline. I'm editing that hook outline relentlessly because I know I have a specific skill set which is making videos. I think I'm pretty good at making videos. And so I'm able to refine it. And so every step I'm able to make it better than the AI would. Just create naturally. And with marketing, it's. Everything is very subjective. And so the first thing that you need to do is just like do the. If you're brand new to some form of marketing or content, maybe you're starting a podcast, like really try and figure out what makes a good podcast. Once you solve that, then you're able to turn it into a repeatable thing. The Unfortunate fact around AI marketing right now, you know, there's a lot of companies releasing products or agent products.
- [09:16]
A
I've looked at every single one of them.
- [09:18]
B
Yes. Basically what these companies are doing is they connect an AI agent to all of your existing tools and, and I'm not kidding, they literally do random things and they do them in such a way that like, it's like a slot machine where it looks like stuff is happening but nothing good is happening. And you'll notice that all these companies who automate all of their social media, if you look at their company's social media, you'll see that they get no views, they're just like spam. And so what you need to do is get really good, use AI to turn it into a process while maintaining how good it is. It's very important not to lose quality because I would never trade quality for quantity right now in marketing. I think the reason you have found my content probably is because I put a lot of effort into my videos and I'm very passionate about what I talk about. And you can't sacrifice quality for quantity. But if you can start creating more stuff at the same level of quality and you just get AI to do all the boring stuff, that's just like golden for marketing, in my opinion.
- [10:19]
A
I think this is a critical, critical point. Kieran and I, we wrote a book called Loop that's coming out on September 22nd. And this is one of the things we talk about, is that to win in today's world, you need quality and quantity. Actually you need to be prolific. You kind of have to be like T. Swift, where you're like, oh, she's making great music and she's making it like every six to 12 months. Right. Versus like every four or five years. And what I think really resonated with me about what you said that I would kind of frame is like, if you are just going to outsource your taste to an AI model, it's going to be sloppy and spammy and nobody's going to care. If you use AI to scale instead of outsource your taste and sensibility, then you're going to be able to create more really great things. And I think you're like a prime example of that. It's like you've built systems to basically get more of your ideas out in the world in a better polished state.
- [11:10]
B
Yes, I 100% agree. And I call this the human center of mass. Take TBPN for example. TVPN created, I think, one of the best niche shows ever Created on the, in this kind of Internet first world. And what they did is they created this show that's incredibly high quality, right? And that's what I call the human center of mass. They don't use AI for anything really. Like they just created this show, they have this team, they have multi camera setups, they interview really high quality guests all the time for like three hours a day, five days a week, right? And so that's their human center of mass. Now they have the power to use AI infinitely to create clips, to create newsletters, content arounding this like very human thing that is live so people will pay attention, right? Because the clips and the automation around the content, that's easy. But if you don't have that thing that's valuable at the center that you're just going to be amplifying nothing. And an amplifier is like something times something. If the first number is zero, no matter what you multiply it by, it's
- [12:16]
A
still going to be zero.
- [12:17]
B
That's kind of how I think about it. And this is why we're going to build a massive set here in New York City. And I'm going to be creating as much content as possible. I'm just going to be constantly learning and I'll be doing a lot more live streaming because once I do that, if I create something valuable, we can just create systems that post clips all over the Internet. The Internet rewards that if you have that thing that's valuable at the center.
- [12:40]
A
When a buyer asks AI for a solution like yours, do you know if you're showing up or not? And HubSpot, we've got a great product that actually helps you get discovered in AI search. HubSpot AEO helps you show up in these moments with the right answers buyers are looking for before the first click, before the first fill. That's the moment HubSpot AEO is built for. Go check out HubSpot AO. You can trial it for free and it's going to help you understand how you're showing up in AI search and how you can show up even more.
- [13:12]
B
So remember earlier I asked for a hook outline and you have two options, right? I can click this and it will open up the notion doc directly or if you right click and hit open in browser, check this out. Codex will open notion. So this is my notion doc.
- [13:29]
A
Oh, that's cool.
- [13:31]
B
And look, it's just like anthropic just brought Fable 5 back online. And this is not a normal model release, right? This is actually how I would probably do something like this. With an intro. I'm very nitpicky, but it at least it like gives me that initial thing where I can just kind of modify it. And I don't like long scripts. I want a single, like a one pager. Like if I had a one page document to do an episode, that's perfect. I don't want any more than that. And so this is kind of exactly how I do it. It does a bunch of research. Codex can do research. One thing you can ask Codex, say, like, hey, I want more hyperlinks in the outline. Just because I want to be able to talk about the primary sources where you got this information. So can you please add those hyperlinks in the outline? Also use many sub agents to do research on different parts of this and then just update the notion doc. And so that's kind of one thing that I would do is because when I make a video, I like to pull up the primary sources. And I noticed here that they don't have a ton of primary sources. And so you can just ask AI to change it. And on top of this, you know, I'm actually noticing here that there's not a lot of hyperlinks. I can just say, by the way, when you're done with that, please update the skills such that whenever you're researching something that requires the Internet, you should always include the hyperlinks inside the outline. I want to see all hyperlinks that you use. You should never write a bullet point that references some research you did and not cite it. So please add that to the. And just to show here we can add it to the hook outline skill. And that's just kind of a dynamic way that you can use skills.
- [15:06]
A
I think, to me, and if I'm the average person watching this, and I'm not doing any of this yet, you now need to work with AI and yourself like you used to work with other people, right? Where you used to send an email and say like, hey, I just had this idea, or I just found this thing that's really important for this project. Can you go and add it to this project? Or I found this data point that's really relevant to this thing we're doing. It's not just about sharing that with another human. It's about sharing it with your agents with your skills so that they have the context and understanding necessary and that you're actually getting like the. Again, I think one of the core themes of the show is the repeatable benefit of actually building the skills and foundations.
- [15:44]
B
Yes. And I think a Lot of people overcomplicate skill creation. I've seen so many videos online. This is how you create skills. There's a template to create skills when first of all, every single platform is different. You know, the way you create skills in the cloud app is different than the Codex app. So the best way to create skills, get the agent to do something, then correct the agent, get it to do something better, and then tell it to say, hey, I really like that. Please turn it into a skill so I can reuse it and then you can give it a name for the skill and it will just create it. And my intuition is that the way skills are created will change over time. It may be completely different than these, like, markdown file system we have now. It may start looking like a skill tree. And so I don't think it's worth it going deep into like how to construct a proper skill, because if you just ask the agent you're talking to, they can create the skill for you. And so that's one thing I'll add.
- [16:35]
A
I have found it's not about creating the perfect skill. It's about having the agent create a skill and then having some basic evaluation step of like, yeah, hey, this is the kind of output I want for this skill. Let's let me test it. And do I actually get the output that I want? And if I do, then the skill's good. And if I don't, I need to be specific about what the gaps are and give the assistant and the model that feedback and it will fix the skill for me and then the skill will be good.
- [16:59]
B
Yes, I think that is the correct way to do it. 100%.
- [17:03]
A
Okay, so what I think is interesting is you're going between skills and the applications and really showing us how you actually build an AI native workflow and work very differently in this case for building content, which is something near and dear to my heart. So let's say we've done all of this. This is also kind of unique and stuck in your codex on your machine. How do you collaborate with other folks that are building your content with you?
- [17:29]
B
Yes. Okay, so this is 100% the next way that we're going to act with AI agents. Claude released something recently. It's called Claude Tag. Cloud Tag is a way to basically take kind of the fun parts of Claude code and turn it into this super agent that you can use in a team. And I think it's really interesting. However, I'm not sold on being vendor locked for this type of thing. And what I mean by that? For those who don't know what vendor locked means, it's. It's just basically saying OpenAI and Anthropic will try and get you to use their agent platform so that you don't use any of the other models. Because on Claude Tag, you can't use GLM 5.2, which is an open model that was created by a Chinese company. It was like actually a really good model and it's a lot cheaper than using their models. And so you want to be always trying to optimize. And for most marketing tasks, as we talked about earlier, won't require frontier models. And so I like to set up agents in a way that I can use any model. And these are agents that are becoming more and more proactive and they operate in Slack. Anyone can talk to them at any time. And no one's figured out this workflow yet. I've talked to 50 people about this. Hundred, actually. A hundred people on how they're setting up with their team. People have figured it out for coding. No one's figured it out for anything else really, like operations or marketing, especially because it's so subjective. And so we're in the early innings of this. So that part's really fun.
- [19:10]
A
Most people are just like, I don't know what to do with agents and if I do, I just kind of do my own little Codex or Claude code. And I think extending them beyond your desktop is a really important thing. And Slack is one way to do that, right? It's not the end all, be, be all way, but it's one way.
- [19:26]
B
This is actually something that we've been really passionate about. And so our company in New York, we've been doing a lot of consulting recently just because I have a big audience and people are coming to me and they're like, hey, Riley, like, how do we set up agents in Slack? And you know, two months ago we just started setting it up for people. I have a really talented team and we just kind of like productized it. This is actually the first time I'm talking about this publicly, that we're kind of going to launch this product in a little bit. And we've been trying to figure out how do we add agents to Slack. And this is a Paul Graham agent that I created. I think I created this last night. And Paul Graham agent is an agent running on a cloud computer. And this cloud computer never dies. So this agent can work 24 7. And it has a specific set of skills and it has a list of connections which are Kind of like plugins, so it can be connected to, like, the company's data. And the same way that when you use Codex, it is just an agent running on your computer, which works for individual use cases. And it also works only when your computer is on.
- [20:31]
A
Correct. These are the two things that drive me nuts. It's like, I feel like I'm all by myself and I feel like a computer has to be on all the time.
- [20:37]
B
And so. But there is some benefits to having it on your own computer. Because it can read the documents on your computer. It can very easily control imessage. Oh, by the way, it added all those hyperlinks, which is really cool. Fun fact, but one of the coolest use cases of an AI agent is I had it analyze six years of my imessage data and just analyze it. And I was pushed to tears multiple times because, like, it was telling me about friends that I've drifted away from since college. It told me that I need to reach out to my family. Do. I'm not kidding. It was one of the most. Like, I was like, oh, my God, I called. Like, I spent like the next three hours after that calling old friends because I hadn't talked to them. And, you know, and life's hard. Like you. You drift out like people you were really close to way back in the day. Like, I haven't talked to them.
- [21:21]
A
Get 6 years of iMessage data exported.
- [21:24]
B
It's just reading them on your Mac. You have to set up a bunch of permissions. Like, yeah, Codex is like, I can do it. But, like, you really want me to go. It's like, do you really want me to have to. You have to enable these three things manually and it can be catastrophic. Blah, blah. I'm like, yes, just do it. And yeah, it just. It did. It took like 40 minutes. Analyzed all of them, it categorized it. So I had to create a little dashboard of like, the different interactions. It was just a crazy experience. I actually did film it. I just couldn't release it because it was so personal to me. Like, there was just too many. I would have had to create code names for all the people. And like, I, you know, I don't want to do that to my friends. So it was just like a truly fun, like, personal experience. Anyway, also with Codex, it has something called computer use. It is the best consumer available technology that can control your computer. Computer use will basically control your computer as you. And this is actually something you should understand as a marketer because a lot of marketing is using software like, canva maybe you're using Photoshop or you're using tools on a computer, AI agents. And this is scary. This is genuinely scary. And I'm scared about this. And I'm on the frontier of AI, or at least I totally close to is AI. Over the next 12 months, will get better and faster at controlling a computer than most humans. Like, almost all humans, like, it'll be able to control a computer. I was talking to someone at a Frontier lab and they were saying they're like, this is a fun time, right? When you use computer use, you kind of watch the mouse move around. It'll stop for a little bit, move around, and it's starting to get faster. But he was saying within the next 12 months, you won't even be able to follow it. You can't even watch it work because it will be moving that fast if you have a fast enough computer. And so I will be able to be faster than you at controlling a computer, which you better get really good at. Giving it instructions. Giving it an agent instructions will become more important. Beyond this, Codex also released a tool called Record and Replay. For this specific reason, what Record and Replay does, you could say, hey, can you please record my screen and then turn it into a skill? I'm going to call this skill Canva Controller. I don't know. Basically, you can take a workflow that you do as a human on your computer. You can ask Codex to record your screen and it will just start recording your screen. Watch. It's going to pull up a screen recording. It takes a sec. I've confirmed the recording workflow. Once capture starts, I should stop and wait for you to tell me you're done. So now for up to 30 minutes, it will cut you off. At 30 minutes, I can just do a task. I don't want to show it now because that'll take a while. But, like, once I hit stop, this is what would happen. I'm done recording. It literally enters this in automatically and the agent will be like, I've stopped the recorder. I'll inspect what you did on the screen recording and turn into a skill called Canva Recorder. That is the future of skill creation. It's not even talking to your agent or getting it to do something and then turning into a skill. It's just recording yourself, showing it. Yes.
- [24:20]
A
And that goes back to your point where, like, as these tools can use your computer more reliably, it makes sense. You're just like, hey, this is how I use my computer on a regular basis. Why don't you know exactly what to do and then I can just tell you so I don't have to go and do it myself.
- [24:35]
B
Exactly. And that's one of the reasons why I think like Mac Minis and hardware, especially Mac hardware, is going to continue to get more expensive if you need to buy a computer. I know the prices just recently increased. I think they're going to go up more. This is not financial advice or this is not investing advice. This is me saying, like, because AI agents can control computers, the demand for having many of these computers will skyrocket. You know, if, if companies can truly get an agent to control a computer and they need physical hardware to do certain tasks, there's no limit to how much companies will spend. They will drop $50 million on computers if it is worthwhile to them, if they see a positive roi. So that's one thing. Back to Slack, because we're just talking about physical computers. And there are some benefits to getting AI agents to control your computer. But most tasks, I would say 95 plus percent plus tasks, can just be done in the cloud and they can be done through clis, they can be done just through basic API calls, because every action that you can do on Notion can be done by an agent now. And they've basically created Notion so that it can be controlled by agents. That's coming to everything.
- [25:45]
A
Right.
- [25:46]
B
And so, yeah, I think you can create these cloud agents and so I can show you how Paul Graham is, is set up. Okay, so this is not publicly launched. We're going to be launching in a couple of weeks and we just want to get some feedback. And so this is kind of how we're setting up our product. Right. Basically you have this agents MD file that the agent reads every single time it runs. And you can actually manually edit this and you can very easily just add connections so you can create a connection here. And so let's just look up Notion. So I'm going to connect this to my Notion. This is my Vibe Code Studio team. Now. The agent has context over my Notion. It's that easy. And then in the overview you can like ention Notion. So you can say like whenever you create advice, put it in a database inside Notion and you can just create this document and this is kind of the source of truth and you'll see that there's like file references. And so you don't want to make your agents MD file too long. You can just add references to where they show up on a computer. This is a computer that's running in the cloud, basically. And it runs 24 7. And you can also set up cron jobs, right?
- [26:58]
A
Yep.
- [26:58]
B
And again, I never even manually go through this process. I just tell the agent to create an automation and that's all cron job. Whenever you see the word cron jobs, you think of automations. And so that is basically how I think about it. It's like you have the connections and then you have platforms. Platforms is the coolest part. You can add it to Slack, you can add it to imessage and you can even put the agent in group chats. We actually got that working last week and it's way more fun than I thought it would be. And so we're moving from a world of single player AI agents, you know, on Codex Cowork to multiplayer. Now this is where companies are running into problems and I'm not speaking just from us. Anyone trying to solve this multiplayer agent problem is are you trying to create a God or are you creating kind of mini employees? Right. Do you want a content writer agent or do you want an AI agent that runs everything on your company with like different skills that you can ention you know that that can just do anything in the company. And we don't know what the right answer is because it would be annoying if you have 30 agents and none of the agents have a shared memory. So like the. Because you might want to use one of the agents for something you're using with another agent. Like if you had a content researcher agent and then the content creator agent, you would definitely want them to like be able to talk to each other or like get memory. And so we're in this very early inning stage of creating agents that anyone in a company can use. But the right way that it's set up is not clear yet and a lot of companies are working on it, so it's pretty fun.
- [28:32]
A
I think what you're showing us is the frontier of collaboration and work and how it's changing. And some of that's uncertain, but some of it is certain in that there are really personal and unique things to each human which you're going to want to do on your local computer. Your imessage example is a great example of that. Right. But then there's a lot of tasks that you need to do with others and you need to do in the tools and products that you work with those people in. And this Slack and this Paul Graham agent is a really good example of that. So if I look back on the show as we kind of close things out, Riley, it is. We have Shown you the importance of recording, documenting and understanding your work so that you can build repeatable skills. We've then showed you how you could even move that kind of agentic experience into new platforms like Slack and better integrate all of the products you're using. As we close out here, is there anything else, Riley, that you think people need to know about building in Codex, building modern agents for their day to day work?
- [29:34]
B
Yeah, I think we covered it from an individual perspective. Right. Getting the agent to do something useful and then turning it into a repeatable workflow, workflow task. Many of you listening, maybe you already do that. Maybe you use skills kind of similar to the way that I was using it. Maybe you're more advanced than I am. The next phase is trying to figure out how to do it with a team. And that's why I think Slack is winning, is because like, you know so many times like our marketing team is now six people and I want them to be able to create a hook outline the same way as me. But with Codex it's not that easy like. And you should also be able to collaborate on it. When if Trevor is asked to do some research on an episode, we should be able to see that document in Slack or somewhere where we can access that and kind of iterate on it together. Right, that's the team collaboration side. And so that's kind of the next phase. Step one is learning how to use agents on your own computer, individual create skills. And then you want to basically figure out a way to like turn that into a team. And there's all bunch of problems that come with this because you have to figure out permissions, which agents can people talk to. What is your agent allowed to do? As soon as you introduce an agent into like an external Slack that you share, it opens up risks of the other people trying to like talk to the agent, get it to release sensitive information. There's a ton of risks. And so these are things that companies are trying to solve right now. And the ROI from creating good team agents where your team already talks is just going to be so incredibly high. You're basically adding intelligence to the place you already do work. And so it is going to be valuable. And so that's kind of the next phase. I hope that answers your question a little bit.
- [31:10]
A
It was. And I think you're giving people tremendous perspective of the frontier of agent building and what is changing and how rapidly everything is changing. And because you're doing that, I want to also make sure everybody go follow Riley. He is creating amazing content again. He is one of my must follows. I'm constantly consuming your stuff and I could not be more excited that you joined us on the show today. Riley, thanks so much.
- [31:32]
B
I appreciate it. Yeah, thank you for having me.
- [31:36]
A
Let me tell you about a show I love hosted by my friend Ross Simmons. Each episode hosts an in depth analysis of some of the greatest creations and creators of all time with deep dive conversations on the creative process that went into building companies, brands, stories and more. If you like learning about history, learning about the creative process, you'll love this podcast. This podcast is perfect for anyone who wants to learn how to systematize their creativity but still be organized. Ross just did a fantastic episode about Reddit and AI distribution in the playbook that everyone is missing. Listen, wherever you get your podcasts, this
- [32:18]
B
data is wrong every freaking time.
- [32:21]
A
Have you heard of HubSpot? HubSpot is a CRM platform where everything is fully integrated.
- [32:26]
B
Whoa. I can see the client's whole history. Calls, support tickets, emails. And here's a task from 3 days ago I totally missed HubSpot grow better.