How I use AI as a creative tool

[00:00:00] Dan: super excited to be here. I've loved collaborating with you. I read Ben's Bites religiously. And it's so good to get to finally meet the community. For people who don't know my name is Dan Shipper. I am the co founder and CEO of Every. We're a daily newsletter. We write long form essays about what comes next in business and technology.

I write specifically about AI. AI as a creative tool, AI as a productivity tool, as a thinking tool that kind of stuff every Friday and a column that I call, that's called chain of thought. And I also have a podcast called AI and I, where I interviewed Ben and a bunch of other people, and we do something similar to this, where we're, sort of going through how interesting people use AI in their work and in their lives with screen sharing and all that kind of stuff. And in addition to all that stuff at Every, we have a bunch of other offerings that we've built around the newsletter. So we have courses. We just we just launched a core course called how to write with AI which is exactly what it sounds like.

And that's like really going very well. I think we have about 80 students registered that's [00:01:00] starting, I think next week. That's really awesome. And then we have a bunch of software tools that we have built around the newsletter and that are bundled into the newsletter subscription.

1 is called spiral. And spiral is it's basically a powerful prompt builder that helps you automate a lot of your repetitive works or repetitive creative works. For example generating the tweet that goes with my podcast, that's like a repetitive task. Okay. And spiral helps me get to 80 percent on those tweets and I'll show you how that works without really having to do much work so I can spend most of my time making it great.

Instead of doing the first draft. We also have a product called sparkle. We like the names at every for some reason. And sparkle is a automatic file cleaner and I can I'll show you like, just really quickly what that looks like. Just keeps your desktop documents downloads or any other folder clean with AI.

And then the last 1 is Lex, which is the spin off app that, ben mentioned he invested in and Lex is a, it's, you can think of it like Google docs, but with AI deeply integrated into every part of it. So you can use AI in context and I'll share [00:02:00] that as well. So we've got a lot of stuff going on.

It's super exciting. I love my job. Like I just get to think about ideas and you use AI all day and play with new toys and stuff like that. So I have a lot to share. And I'll, maybe I'll leave it there. Ben, any, anything you want to talk about before we dive in?

[00:02:15] Ben: No, I think it's just interesting to see how.

Like you started what seemingly just felt like a newsletter to then all these creative things spinning out of from what happened at work. It's Oh, actually this could be something and this could be something. And I think it's more of a new, it's a new, it's a newer thing for you looking at. The kind of business every is right.

I think you wrote about that recently saying it's not just like publishing content. Now we also are spinning up all these other things. Do you think that's because AI is where it's at and it can enable you to build stuff quickly? And how do you, how are you looking at that?

[00:02:55] Dan: It's a great question. Do think that there are other creators or creator type businesses [00:03:00] that have done this model prior to the AI wave.

Example would be someone like Sam Harris, who has a podcast and a newsletter and he writes books. And then he also has the waking up app, which is a really good meditation app. It's headspace. You can look at also the New York times they have the newsroom, they have the cooking app, they have games and portal.

And so I think it's like a. It's a path that has people have gone down before, but it's not like the sort of standard venture path. And I think it's becoming more and more a thing. I think you're doing something like this as well. And and I think I is an opportunity in 2 senses. 1 is there's just a.

There's just a ton of white space for taking this technology and applying it to different problems. So it's a, there's a whole new generation of like opportunity basically for things to build. And to, like you said it just makes it much cheaper and faster to build stuff. We're a really small team.

We've got seven full time people. We only raised like 700 K ish, and we raised almost [00:04:00] all of that in 2020 and never spent it. And we've. Built three different apps. We've got all these different business lines, all that kind of stuff. And I think our pace has been able to be like way higher because everyone on the team uses these tools really religiously.

[00:04:15] Ben: Yeah. I think that the pace and also something which I used to hop on about with no code was that if you can build it quickly, you're less emotionally attached to something. So if it doesn't work, you can close it down and you don't feel like you're pushing that rock up the hill. It feels like that with AI, like these things.

They almost don't need to make sense, business sense for you to do them, but they might like, they might be when they're in the hands of someone like that. The Sparkle app, for example, to me feels like there's a cool tool, like that doesn't become a billion dollar business on its own. Yeah, totally. But you wouldn't do that necessarily if you're building a business traditionally without AI,

[00:04:56] Dan: oh, you're 100 percent right. And I'll show you Sparkle just so you guys can get a [00:05:00] sense this is what it does. So this is what my desktop looks like, my down, can you see?

[00:05:04] Ben: Yeah, I was gonna say it was just on your Discord then, so yeah.

[00:05:07] Dan: Okay. So it's, we've got my this is what my desktop looks like.

This is what my downloads look like. It's all just like super, super clean and I don't have to do any of it. It just happens automatically. And so we've got an AI library. This is where the AI will like automatically organized all the stuff that it finds on my desktop. We've got recent, so this is all the stuff that has recently been added to the desktop.

Same thing for downloads. So it's easy to get stuff I need to attach. And that a manual library where I can just organize things manually, anything that I just want to do myself and you're right, like this. Honestly, this is an app that like Apple should build into their operating system and they probably will.

But in a world where we can build this like really quickly for very little money it makes a lot of sense to do. Cause it's like, who knows where this is going to go. And for me, like personally. It's just such a breath of fresh air to have this running all the time. And I have direct access [00:06:00] to an audience that might also like that.

And so it starts to make sense to do things like this. And I think it's I think it's really good for a business. Just. To provide more value into the bundle that we have, it creates a lot of stuff for us to write about because we can write about the experiences that we've gained from doing this.

And, as like the more you can experiment with trying different things and building different products, the more likely you are to find something that really works. And so our iteration speed is much faster because the bar for trying anything is much lower. We can just go and do something in two days that becomes something interesting.

[00:06:34] Ben: Yeah. And I think it's your your brand has taste is what I said to you before. And there's a voice and there's like people behind it. So there's a lot of you in the business. And I think, like you said, when you're building stuff, that's this would be helpful for me, your audience is lots of people who are like you, there's lots of crossover.

So these creative businesses seem Yeah, really interesting places to add on these extra tools that are like small [00:07:00] utility tools that just add a bunch of extra value. Like you said yeah, I think that's awesome. Totally cool. Should we should we dive into some workflows and some other stuff?

[00:07:09] Dan: Yeah, let's dive into it. I think the first place to start, I'm going to share my screen again. Let me know if you can see, I actually, I got to grab a book. Yeah, I think the first place to start. The first place that I start is if you're making stuff the output that you can create is a function of your inputs.

And for me, a lot of my inputs are reading and a lot of a lot of the stuff I read is actually like difficult to read. It's like a lot of. Like philosophy stuff or like literature or computer science stuff or whatever. And I find that chat Bt Claude too. But Chatt in particular is like really good for this.

And I actually like like I knew I was presenting this today, but like I can show you just like actually what I used this for. 'cause I was just using it literally this morning, just organically for this exact purpose. I'm reading Moby Dick. [00:08:00] Which great book. I've never actually read it in school.

It's like weird and it seems totally irrelevant to like AI, but I actually think that the likes the 1850s period in America was this like, new period in America where there's this new paradigm of thinkers like. Melville and Emerson and what Whitman and I think you can take, I think you can directly trace a lot of the thinking in that period of time to the kind of mindset that created language models.

And I'm writing something right now about that. And so I'm going back to this like source text and just reading Melville one because it's fun. And two, cause I actually think it has like a direct bearing on my work. And what I do is like I literally just take Cheshire BT. I have access to the new advanced voice mode feature.

So I can, I'll bring it up here. And what I do is I literally just, as I'm reading, I'll ask it questions. So I'll, there's a, there was a line in this, in the book that [00:09:00] likened a character to the difference between rot nails, W R O G H T and cut nails, which is, I don't know. I just wanted to understand that better.

And I just had Chachi Petit open and voice played listening. And I was just like, what is this? What does this mean? And that might feel like a kind of irrelevant or pedantic example, but I think what's really interesting is having Chachibit open and just being able to ask it questions as I'm reading has revealed to me, like, how much, how many of those questions I have.

And how much more I can get from the things I read, whether that's fiction or like computer science stuff, or just like anything in the world, just by having ready access to something that will answer those questions. And so I just get I get way more out of my reading by either having it just listening to me as I read, or just asking a question as I'm reading just by texting it.

So like another example is [00:10:00] I'm reading another book which overlaps with with Moby Dick, which is called All Things Shining by this guy Hubert Dreyfuss, who's a philosopher who's written a lot about artificial intelligence. He's really actually very cool. And this particular book touches on Moby Dick.

And I just really wanted to know, and I think this is a common thing, like he makes an argument about Moby Dick and its philosophical implications or whatever. I just really wanted to know What do people think about this? And again that might seem like a very specific literary who cares what people think about this argument, but I think that happens all the time.

If you're reading blog posts, or news articles, or whatever you want to get a sense of the Prevailing opinion and whether this person what are the, what is their argument? What are the counter arguments? Is this person right or wrong? And if you're trying to develop a view of the world, which I think is really important to do any kind of creative work, it's really helpful to have or Claude to help explain to you what the prevailing opinion is, what the most up to date information is, what the counter arguments are, all that kind of stuff.

So that's one thing is I think it's incredibly [00:11:00] good tool for reading, for deep reading. I have tons of other use cases that are that are reading related, like a really good one is just take a picture of the book, ask it to rewrite it or translate it for you say, give you the passage and really basic terms.

And, if you're reading technical stuff, like that's really helpful. Another one is just And the passage to Dolly and ask it to draw a picture of it. And just getting it in another modality, I think, often really helps you understand it. So that's that's the 1st 1 is reading.

I'll pause there in case you have any questions. Ben, if you have any questions, if anyone on in the chat has any questions, happy to expand on it, or I can move on to the next 1.

[00:11:33] Ben: Yeah, so actually this is something I tweeted something yesterday, and I didn't tweet it very well But I was thinking so I started reading a few books recently that I have struggled to get into And often it's because I'm reading late at night when I'm just about to go to sleep.

So as soon as I'm going to sleep, reading is the last thing I've been doing and maybe I've trailed off towards the end. So like jumping in at the middle of chapter 13 or something and going back and [00:12:00] being like, Okay, this is where I left yesterday. I can't remember where what was going on up until this point Like I was thinking having something that is Just give me a like a episode recap like you have on netflix shows Just like what's the recap of this so far to like quickly get me up to speed.

I think that was That would be an interesting way to use AI, and I'm sure there's different ways to do that. But I liked what you said, where you're looking at what people are saying about this. Because if you're reading something that you like, or there's a there's some thoughts coming from you from reading that book, passage, whatever, especially if there's an old book, There wasn't Twitter and blogs and stuff around, necessarily, when these books were written, where people could write their own thought pieces on their interpretation of the book.

And it's quite interesting to be able to see what were the different opposing views of the thinking in this scenario, versus when, like, when you're looking at it, be able to see the different, how different [00:13:00] people are reacting to that, Writing, right?

[00:13:02] Dan: Totally. I think it's like an intellectually curious person tragedy between Claude and this generation of models or stuff.

It's like being a kid in the candy store because getting, If you like reading and you want to understand things like in order to get any kind of like good interpretation of Melville I would have to go to grad school, which I definitely don't want to do. And this sort of lets you get a grad school type level of depth on any topic, whatever you're interested in.

And for me, and I think this you, Ben, I just love going down rabbit holes. I love getting curious about something and just learning everything I possibly can about it. Sometimes it's literary, sometimes it's philosophical, sometimes it's business, sometimes it's just something like really like much more practical.

It's growth marketing or whatever. And I think whatever rabbit hole you go down, this is it's like a super tool for that. And it, it gives you the kind of information or the kind of like intellectual companion that you would only get if you went to grad school or you're in a particular kind of environment that I think a lot of people like just [00:14:00] can't be in for whatever reason.

So yeah, I think it's great.

[00:14:03] Ben: Yeah. Second is asked, do you find it can help you analyze the structure of what you're reading to apply it to your own writing from articles to essays to even if you're going to do your own book?

[00:14:15] Dan: Totally. And that was actually going to be my next thing to talk about is I think one of the, one of the big, Benefits of AI aside from like once you've done the kind of like input you're like you're starting to get things into your head.

You're starting to go down rabbit holes or whatever. I think another really important part of creative work is being able to say what you like and why what is your taste? And I think that's really important for several reasons. One is if you can identify your taste you can start to create a target for the kind of thing you want to make.

You can say this is good. This is not good. This is like what I'm trying to do. I really like these writers. I really like these movies. I really like, these products. And you can articulate like what [00:15:00] about those things Lights you up, and that becomes a target for you. Another thing that's really helpful and important about this kind of thing is it can help you bump into more things that you might like, and that becomes more fuel on the fire for you to do more stuff.

So a really good example, just, also from today is let me show my screen again. Give me one second.

is I think I did this. Oh, yeah. So when I was asking it about, this guy, Dreyfus is a philosopher. He wrote a blog on Moby Dick. I really like his interpretation of Moby Dick. And I was just like, where does this come from? And it's like Martin Heidegger, another philosopher, Friedrich Nietzsche, another philosopher, William James, who's like a American pragmatist who I love.

Yeah. And then it names about a bunch of other scholars and it's okay, like I'm into this. Now I have a whole reading list that I can just like keep going down the path. And this would not like, I don't know how I would have gotten this before. I would have had to go to some Moby [00:16:00] Dick scholar and be like, Hey I'm an internet writer.

I'm like, interested in AI. What do you think? What should I read? And that would have been like, fuck off. I don't, I don't know. Who are you? And this This is all available. And I think previously, like with Google, in order to get this list, someone would have had to think about creating an article like this.

And this just writes the article for you based on the sort of like vibe of what you're going for. And I think that I use this all the time for similar kinds of tasks. So for me, one of the like, one of the I wrote about this in an article called what I do when I can't sleep.

One of the like big unlock moments for me AI wise is like being able to name my taste. And so one of the things I did a while, a long time ago, maybe about a year ago is I just took a bunch of literary influences and I threw it into a chat these are all authors I like and I was like, here's a list of authors.

I like and you can do this for movies. You can do it for products. You can do it for like pretty much anything that you're think about anything that you're trying to make and think about people who have done things that are like [00:17:00] what you want to make. And if you don't have that list, you can ask to help you with it.

So I put in a bunch of people. We've got William James. We got Mary Oliver. These are they span philosophy, literature, science business music, a bunch of different places. And I was like, can you tell me the vibes of these writers? And what Chachaputi is really good at is like doing this kind of like summary where it puts its finger on things that you probably have felt, but have never been able to say before.

And that's an incredibly important thing for any kind of creative work. So like an example would be like William James, like philosophical warmth. I've never actually done this exercise a lot. And I've never said that before, but I love that. That's there's some There's something about that.

That is so core to what we're trying to do with every, that I I always try to articulate and it's something about like curiosity or like open mindedness, but the idea of warmth and philosophical warmth is like, it's like a really good way to phrase it and that becomes once it's named that becomes like a [00:18:00] target that I can point myself out or I can point like people at every app and be like, this is what we're trying to achieve.

And it's just really good at naming things like that, like world building or like wonder or. Depth, playful depth is really interesting. And then I can be like, okay, cool. You synthesize these together into a cohesive summary of what I like. And it'll give me like a pretty good, more compressed summary of what used to feel like pretty disconnected into something where it's okay I'm looking in a mirror.

I get what I like. And it's okay, I'm drawn to writers who blend intellectual rigor with emotional depth, often exploring big existential questions through accessible human centric prose These authors often stand at the intersection of curiosity and reverence. There's a strong pull towards vulnerability and emotional honesty.

Ultimately, you like writers to help you understand both the inner workings of the mind and heart and the outer complexities of the world, all while maintaining a sense of wonder, humor, and human [00:19:00] connection. And I'm just like, damn yeah, that's fucking but I I could never say that before.

And and I did this exercise like about a year ago and It really helped me figure out how to articulate what I wanted to do with my own writing and where I wanted to take it every and honestly since I did that, it has been like it has the whole business has just exploded in this really interesting way.

And so I definitely I definitely think exploring your taste with these tools, whether it's in books or movies or products or businesses or like anything is like incredibly important.

[00:19:34] Ben: Yeah. So interesting to see what comes out of that. Like even just reading that, like if I was given a question saying where on the internet can you write, can you read that fits with this description, I would be thinking of every, of That's a perfect description of what every and the writing that comes from every feels like to me, and it's all here.

And it's funny, obviously, that you actually did this exercise, which is [00:20:00] where then some of that direction came from. This is related actually to a question that Jason's asked, which is, Do you then use things like this for robust custom instructions or memory, like any suggestions or tips and stuff you have around going through some of these exercises, then actually using that as like totally working partner with chat.

I

[00:20:22] Dan: definitely do. So let me see if I have custom instructions on this version. I sometimes turn them off cause they do a lot of demos and it gets weird. Are you, they keep changing their interface around. I don't know. Here we go.

Let me look this up for you. So I have two things that are like this. One is in Chachibiti, and one is in Claude. And I, give it a little bit of background on who I am. I give [00:21:00] it a little bit of background on what I love. And here it is. I love evocative, lyrical, accessible, emotionally vulnerable, analytical, intellectual, philosophical, accessible, writing with dry humor.

This is literally just from Chachibiti. Chachibiti. And it telling me a little bit about like myself and my goals and my psychology and who different people in my life are. And, it, and then I tell it things about like things I know about myself that I want it to help me with.

So I know I'm like a little too agreeable. And I know that I'm a little bit too opportunistic. Like I just get too excited about something new and then I'm like, let's do it. And I think that's really helpful when it's just you or whatever. But when you have a team, it can feel a little jarring and I need to hold myself back a little bit.

And I will often before I'm. As I'm making a big decision, I'll often go to Chattopadhyay or Claude and ask it to help me unpack the decision. And it'll be like, if I'm, dealing with some interpersonal situation, it'll be like you're often too agreeable. What would you say in this situation if you weren't afraid of feeling [00:22:00] guilty or feeling like bad for saying no to someone?

And just that little question just helps me like filter out some of the agreeableness so that I can make a decision that's like a little bit more clear headed. And so I think it's a really good tool for managing your psychology in that way, where you're often like, trying to take an insight about yourself and apply it in your day to day and chat.

You can call it a really good reminding you of what those insights are. I have another, thing I have a cloud project called my psychology, which is like a very similar very similar thing. And really good for a lot of different things. Here's a fun example. We are hiring more entrepreneur residences for and hopefully please don't share this, but we're hiring more entrepreneur residences for our for every, and I.

Asked it, I just uploaded a list, like a hundred people applied. And it was like really hard for me to figure out who I wanted to like even interview. And I just, I have this cloud [00:23:00] project called my psychology, which is just like, it has a bunch of journal entries and goals and writing and like a bunch of stuff, so it knows a lot about me.

And then I was just like, okay, here's this list of 150 people. Who do you think I would like? And it just picked out people who I did end up liking. And I don't I don't know who we're going to work with or whatever. But it's it's so useful to have something that just knows you that you can just be like, okay or another one is there's this movie, Spirited Away, that like everyone in tech really likes, and I just don't know, I don't really like it for whatever reason, and I was just like, can you just explain to me what I might like about it? And it knows me well enough to be like, here's probably what you would like if you were going to like it.

And so yeah, I think having it as that sort of like mirror, not just of your taste, but like in general, like what your psychology is like. Is really helpful for making better decisions doing better creative work like all that kind of stuff.

I think you're on mute. [00:24:00] Yeah.

[00:24:00] Ben: Yeah. I was going to say, I think I need to do this with June. Like I know everyone in tech loves this, these movies and I'm like, I don't get it. I'm bringing that up now, but yeah, I need to do this. Yeah, this is awesome. I love this. Love seeing all of this and like how much you actually go into what you like and why things that you read or consume should relate to what you like or why don't they resonate with you and all that kind of stuff.

Yeah, cool. Let's should we go on to the next thing?

[00:24:28] Dan: Let's do it. Another one that's really good for if you're, if anyone is doing like a lot of long form writing, like I have a piece that I'm trying to write called seeing like a language model, which is, it's back to that like Melville philosophical roots of language models, whatever.

And it's just really hard to write a piece like this because it's so big and dense and it like covers I started in like ancient Greece and I'm. It's just like a whole history of ideas. And I like have this notes file and I've literally just been like adding things to this notes file for a long, long, [00:25:00] and I think this is probably familiar for anyone.

It doesn't matter if you. If you're writing a long piece or you're like, you have a business idea or whatever, like you're often just you have this big longterm project that is changing the way you see the world and you're constantly finding examples of Oh, that relates to this thing or whatever, and you may not have even articulated to yourself.

What is it I'm even really trying to say or do, or what's the problem here. But you have this like intuitive underlying sense of this is relevant or this thing is like important for the, for what I'm trying to do. And so for me I just add stuff to that file and then I upload it into a into a cloud project.

And then I'm like, constantly just being like, hey, what's my thesis? Like every day I'll just ask it like okay. I'm trying to express something like this. You have all my notes. Can you just like really say what I'm trying to do? And it'll give me like a thesis, like the rise of language models represents more than just a technological leap.

It embodies a fundamental shift for understanding of knowledge, intelligence, and human potential. [00:26:00] And then it challenges the morning old platonic quest for universal definitions, like blah, blah, blah, whatever. And this is like very philosophical and jargony here, whatever. But it can help me take the mess of things that I'm intuitively grasping towards and then help me put a little bit more structure around it to understand what I think.

And I'm not expecting it to come up with a thesis statement for me. I'm just saying. Here's a bunch of garbage that I think is related. Can you just tell me what and then I'm going and refining it down and like rewriting it myself. And it's this like back and forth collaborative process where we're I'm using the language model to help me like move closer and closer to what I'm actually trying to say.

So I can actually like write the piece. And I, and so I think it's like really helpful for that kind of that kind of Articulation of big ideas that would otherwise be hard to even wrap your head around. And that, it goes for any, anything you're trying to do or any field that you're in.

[00:26:53] Ben: Yeah, it feels like that is That's effectively the core way that people should be using AI is [00:27:00] start with something big and then refine it down to something smaller that you then have something to put your own spin on rather than that smaller thing is the thing that you end up putting out into the world.

It's more of a Yeah, that person who is like sat next to you and helping you refine and synthesize all this knowledge and be like, no, remember you said this thing like the other time. And actually, this is what you're trying to say. But now you're saying something different. And

[00:27:26] Dan: yeah, totally. And same thing for like meetings, like you have a bunch of people in a room, like all trying to figure out what to do, like synthesizing down what are the different Things that we're thinking and feeling so that we can get a common understanding of a problem and then figure out what to do about it.

It's, it's really helpful for all that kind of stuff. And I think it, it it moves into the writing. I do a lot of the writing for this in Lex, which is like I said, the Google Docs with AI built in that. That we incubated at Every and spun out and it's run by my Every co founder, Nathan Bishhez.

[00:28:00] And for me It's basically just like hooking up Claude to the writing, and it's really helpful because I can go in here, and I can just chat with it, and it has access to my draft, I don't have to copy and paste anything, and you can see I don't know what have I used it for for example, again, back to this relationship between like American pragmatism and this literary tradition in the 1850s to like language models now and I need to write that, I need to write a section of the piece like that and I'm just like, can you just write it, write the summary for me that like I know in my head, like I can't fully get out and then I just get a little bit of a summary of the the main ideas and the sort of the path, like it's going through a forest and there's something ahead of you just like Whacking through the thicket and creating a little bit of a path so that then I can make it nice and do the thing that I want to do.

So it's really nice to have it in context. It's really nice to be able to be like, you can also do this comment thing, or you can do this all the time. And I can be like, what do you think of this [00:29:00] paragraph? And it's just a very clean interface that includes. And it has a bunch of other features that I sometimes use.

Like it has this like focused work set, focused work thing, which is not really AI, but it's just you can time your writing and measure your productivity. It has this thing called checks where you can ask it for and you find all instances of like my passive voice or passive voice, and it will just search the whole document for a passive voice.

And I think I did only the highlighted thing. I was in trouble.

[00:29:29] Ben: Yeah, that this was one of those game changer features for me. So I was using Lex. When we were writing up these case studies, we'd have transcriptions or a bunch of information from different blog posts, put it all into Lex and basically then do similar to what you did and be like, outline this in like the right way.

And I think there's now you can set a template now. So it's this is what our case studies would look like is this kind of template, this kind of format. And then the checks for me [00:30:00] were really helpful because half the time I was like, actually, I am trying to write in that way. So that's fine.

But the other times it would be like, Just telling you how to like the brevity check is you can say this more clearly, or there's just like a few different texts or run a bunch of those. And I think you can do it based off. Isn't there a setting or there's some settings to say this is who my target audience is, and this is who I'm trying to write for.

And this is like the kind of stuff that they're like, and it would do checks based off of that.

[00:30:26] Dan: Exactly. It's like a, it's like you can do custom instructions for the piece and tell it like. This is the context in which I'm trying to write. So read it from that perspective and give me feedback from that perspective, which I really, I really like it.

Just, I think it makes the whole process of writing. It just integrates it more fluidly with AI. I think the mobile experience is much better than Google docs. So all that stuff is really great.

[00:30:49] Ben: Yeah. Actually, could you walk us through, Xenia has asked about this, which is like what does your tool flow look like when you're writing something?

So is it like. I start by dumping stuff into ChatGPT and then I'll go [00:31:00] to Claw Projects and then I get to a point where I'm in Lex or is it really ChatGPT?

[00:31:06] Dan: I would say generally it's so I'll often like, go take a walk either with chat to voice mode or like with voice memos, and I'll just like brain dump stuff.

And that's a really good way to get out whatever's in my head I'll often then transcribe it and throw it in chat to your cloud and be like, what's interesting or help me like, figure out what I'm what's on my mind. Yeah. I think I then spend a lot of time just like writing random notes in my notes app on iOS and in things like I use things where I just like, oh, I write every Friday.

So on Thursday, I sit down and write a piece and end. And usually I have the headline for the piece already in my note in my things. And it's just here's the headline. Okay, cool. I'm just gonna go write it. I tend to do most of my writing in Lex. I tend to use Claude for a lot of the back and forth.

I [00:32:00] think Claude's better for writing generally. ChatGPT is just a better mobile experience. And yeah, it just has a little, a few nicer bells and whistles UX wise. We do most of the editing in Google docs. That's just like the thing that our editors are most comfortable with. Our editor in chief, Kate will often take the thing out of.

Lex and put it into Google Doc. We'll go back and forth with her and our other editors. And then the last I think important part of my writing process is actually promoting it. And for that, we built this tool called Spyro.

And Spiral is basically like I said, it's like this prompt builder that gets you to 80 percent on a lot of the repetitive creative work that you have to do. So like for me a lot of the, a lot of the work that I do is okay, I write an article and then I need to turn it into a tweet.

And so we have all these spirals in our team where it's like let's see, what's one that I use a lot. Okay. I need a title suggestion, right? [00:33:00] So I can take a piece and let me just grab a random piece from every here's one that I just wrote that one really, that one kind of viral. So I can just take the piece.

And copy it and paste it into here and press generate. And it will just use AI to create a bunch of headlines. Why generalists may thrive in the age of AI or generalists that are equipped for an AI driven future, the unexpected advantages of being a jack of all trades in an AI world, I actually think these are better subheads than headlines, but you can just go back and forth and and the way that this all works is if you want to create a spiral, I can say okay, I want to generate I want to take articles and turn it into tweets.

Something like that. And then you give it a bunch of examples. So you can say, here's an example of an article that I turned into a tweet that was like, good. Or here's, or you can just give it output. So you can be like, here's a bunch of tweets that I think are good. And then it will, you press extract patterns and it will write a little guide for itself.

And then you get that spiral where, [00:34:00] anytime you want to turn an article into a tweet, you paste the article and it will turn it into a tweet for you and it just saves a lot of time. Like I use this all the time for my podcast, for example. Should be able to somewhere. So like tweets from podcast transcripts, like I'll literally just this is how I like.

I interviewed Dwarkesh Patel like four weeks ago or something, and this is how I generated the tweet for it. I just took the podcast transcript, pasted it in here, and it like, created the tweet. And I tweeted something that looks very much like this, and it went viral. This happens all the time. I do this also sometimes, like when there's any sort of news event, like I'll just take the news event, throw it into spiral, get a little tweet that summarizes it for people and and post it.

And it's just a really good, and I'm always editing it, make sure, making sure that's my voice and all that kind of stuff. But I think. One of the really interesting things about AI is it reveals how much repetitive work there is even in creative work. And it's really good at, if you give it a bunch of examples of previous things that you've done, like eliminating a lot of the [00:35:00] repetitive aspects of creative work, so you can focus on the actual, like new stuff.

And this is just so helpful. Like spirals are so helpful to have with my team. Cause like now they can tweet in a way that like is either in my voice or in a style that I think is like. Going to work. And so it helps you save a bunch of time and then standardized or push a sense of taste and skill out to a bunch of people in an organization.

More or less automatically and yeah, we built this internally every I built the 1st version a couple days. We had a we have an in house developer who helped take it the rest of the way. And it's grown really well. We use it all the time internally. I love this product. I think it's really cool.

[00:35:37] Ben: Yeah, this, the way it's set up effectively is writing a really good prompt for you. So it's doing those examples, which effectively just building that prompt in a UI that then goes off to Claude, so why, what was the benefit of building this as a standalone tool like this, rather than obviously people could do a similar thing with custom instructions, things like that.

[00:35:58] Dan: I think. For [00:36:00] me, okay, a couple things. One is just it's a pain in the ass to do that prompt all the time. It just is, especially if you have a bunch of different use cases. Second thing is, I might have done that prompt, and I was, like, the reason we did this is because I was doing that prompt, and I was like, this is a pain in the ass for me.

If I'm thinking it's a pain in the ass definitely my team's gonna think it's a pain in the ass, and my people on my team are not gonna do it. And I think this is just more or less like my general theory of one, one set of opportunities and in the AI world is, I think a really good way to think about possibly the next, 10 years of like software business opportunities is to look at How Excel worked and, in the PC age, so Excel was like the 1st killer app for PCs.

It was like the thing that convinced businesses and they needed to have computers. And it really taught a whole generation of computer users like. And a particular paradigm of computing and it's, I think it's like a really unique and interesting piece of software because anyone can use it like, [00:37:00] like you can get started.

It's like super basic to start. Anyone can type numbers into a cell but you can get as complicated as you want. It's just like this unlimited set of powers and it's very general purpose. So it's built for, financial analysis and stuff like that, but people use it for everything.

And that the general purpose nature of. Excel meant that as it taught people to use computers, they started using it for things that it wasn't originally built for. And that became the opportunity for B2B SaaS. So a lot of B2B SaaS companies wouldn't really exist without Excel first having existed to teach people how to use computers in this way and teach people that computers could be used in this way.

And I think there's a really and so Excel ended up getting unbundled with a lot of the B2B SaaS companies that you see today. And I think there's a very similar, probably trajectory for a lot of these chat apps. There's a lot of overlaps where Cloud and TouchBT, they're like incredible general purpose tools.

Anyone can get started with them really [00:38:00] easily. Their power is they're really powerful if you know how to use them right. And as people learn. These new workflows are teaching people a new way of using computers. Like language models are just like a fundamentally different computing paradigm than we've ever really encountered before.

So as people are learning to use computers in this way they're going to create very specific workflows for them. So for me, it's I need to make it, make a tweet out of this podcast. So I have a very specific workflow and a specific set of needs that like Claude can do a hundred percent. But it's just not really built for that because it's, it has to be so general.

It has to it has to accommodate every use case. And I think what Claude and Chachibiti are doing is they, by making AI accessible to the broader public, they're creating a set of opportunities For people to realize that there are problems that they didn't have before. Like before no one was like, I don't it's hard for me to generate, like you, the prompt to generate my, like my podcasts, my tweet for my podcast, because no one even knew that was [00:39:00] possible computers.

But now as adoption of these tools gets, of these general purpose tools gets higher and higher it creates this problem. Place where people are like, okay, this works, but it could be better. And I want something like specifically for my particular use case. And I think that's the opportunity for spiral is Claude is like, and Chad, you're never going to do a feature, never going to do features specifically for Marketers who need to repurpose content, they will always be able to do that.

But I think if you if you build a purpose built tool you'll be able to capture a lot of those people that feel underserved by the general purpose tool.

[00:39:37] Ben: Yeah, it's like that rapper argument where There's two sides to it, and I've gone back and forth on this where the wrapper is just a very specific use case of that thing that it can do, but isn't it better for a specific type of user to just have an interface that does that thing and takes all the complexity out of it?

And that's why wrappers [00:40:00] do so well and they're making so much money. There's a lot of them, like they're not difficult to make either. because they're fundamentally built around the same technology that powers the general purpose tool. I think the most, the biggest companies as wrappers also have to have all these other baked in workflows and knowledge stuff that is probably more traditionally the sort of B2B, SAS workflows and like structure and things like that there.

But it's not to say that individual wrappers are not useful because. They're only useful if people get use out of them, right? So like this exact scenario.

[00:40:37] Dan: Totally. Yeah I there was a whole like rapper debate. I'm just like, I don't know I'll take like this is working. Like I'll take the money other people are worried about rappers like Yeah, I just

I think that we will Rappers are just like the sort of [00:41:00] like Infant gestational stage of what will be, like, very fully featured applications over time and being able to differentiate just at the user experience layer instead of at the fundamental underlying model layer. I think it's actually hugely important especially as, foundation models, like more or less, like they're getting to a place where they're interchangeable and more or less like their level of power is enough for most people. I think that most people are not pushing Chachi, BT and Claude to its limits right now, because they haven't even really learned what it can do.

And so further improvements to foundation models are going to be really important. But they're mostly going to affect for now, like people like me and you, or we're just like in this all day. It's our job to be like at the edge of things and certain people with certain use cases.

But like the vast majority of people, are not even getting the most out of the current layer of [00:42:00] technology. Creating user interfaces that make it more accessible and understandable, I think is like a really powerful benefit.

[00:42:06] Ben: Yeah, and I think the actual debate actually was started by VCs really saying these aren't billion dollar companies, which is a completely different debate for the everyday person who's this tool is useful for me to do this thing.

So yeah. I'm going to use it. I'm going to pay for it. So yeah, we'll jump into a few questions now. We've got a few minutes left. We have Senya, who is the first to raise a hand. I know she's been very interested in this chat. So yeah, we'll come to you to ask a question.

[00:42:38] Guest: Okay. Sorry. I apologize for sledding. This is because I've written can you guys hear me? Okay. Yeah. Yeah. All good. Yeah. I've written a whole bunch of books and my question to you is familiar with the hero's journey. And 12 acts in a screenplay and seven steps. Okay, cool. There's the three columns.

There's all these [00:43:00] structures. So I'm going to develop my own. It's currently there, which is sometimes people would look at it. This is what this looks like a math formula. But I might think, and it's everything from a sentence to have, the most important word of the last piece of the sentence to the whole structure.

So I'm wondering. Speaking out with you and everybody else, your writing creatively, just, I was itching to use the tools to create almost like my own formula because I'm in the middle of writing a thriller. It's been three years since I touched it, but now I'm in a tech job. But have you used that?

Do you have any ideas? Are you looking at exploring it? And you think it's a crazy idea or is this something legitimately to use the tools? I don't mean for it to write for me, but to use it as a accelerator for me to, finish my thriller, for example.

[00:43:49] Dan: Yeah, I think it definitely is.

You should look at there's this company called pseudo S U D O pseudo right. com. And they do AI writing tools specifically for fiction writing. I think it's like. [00:44:00] Incredibly useful. There's tons of different, features they have that are like, okay help me figure out how to write this dialogue or help me brainstorm on different character attributes or like, how a character would behave or help me write more vivid descriptions and show instead of tell there's tons and tons of stuff like that.

I think you can use pseudorite. I think you can use something like Lex. I think you can use Claude. I think from a structural perspective I think it's really. In general, it's a really good thing, like, whenever I read a nonfiction book that's here's a framework it can be the hero's journey, or here's a marketing framework, or here's like a strategy framework, or whatever, I can just throw that book into Cloud and be like, okay, I want you to like, summarize what the framework is, and then help me apply it to this specific case.

It's really good for that. So I think generally it's a super powerful tool and it's worth doing. I have not honestly done a lot of fiction writing with it, but that's just because that's not what I'm focused on right now, but I, at some point I will be back to fiction writing probably, and I will definitely use it.

[00:44:59] Guest: So [00:45:00] you feed the entire books?

[00:45:08] Ben: Awesome. Anyone else with any final questions? Raise your hand or forever hold your peace. I want to know like what, what's exciting you next, like you go down all these rabbit holes in a very deep way. What's one that you haven't quite gone through yet or haven't, you want to go down the one a bit of time to do

[00:45:30] Dan: rabbit hole.

I want to go down. I don't know. Let me look at my notebook and see if I have anything worth worth saying. I've been thinking a lot about like one of the things that we're incubating right now is It's a little app that responds to all your emails with chat with cloud and it's working. It's pretty cool. And so I've been thinking a lot about that and a lot about emails and some of the things I want are like, I think you can apply a lot of this summarization and [00:46:00] synthesis and all that kind of stuff to like companies.

Do I just want right now, a lot of things we do we use this tool called granola, which records meetings. And it's really cool because we have this guy, Brandon, who's he runs our studio and our consulting arm, and he's a good friend of mine. Really smart. And, he's, for example, running the hiring process that we were talking about earlier.

And when he talks to someone I often want to know what was that conversation like? And rather than ask him specific questions I have access to a transcript that I can chat with, and I can just hear about what that was like. And Yeah. I, but then I forget sometimes go chat with transcripts.

So I just want like a daily summary of here's all the meetings that happened at every, and here's all the things that got talked about, and here's what I think is relevant for you. I think so that's one side of it. Another side of it is in terms of repetitive stuff that you do inside of companies I think I'm constantly giving very similar feedback founders repeat themselves all the time.

And I just want a little thing that like gives first pass feedback in [00:47:00] a way that I probably would give. 'cause I think it would just like scale my attention a lot. And so anyway, those are things that are on my mind. Maybe we'll work on them. Maybe we're maybe we won't, but and feel free to steal if you want.

But I think it's, I think those are some of the opportunities of that are on my mind right now.

[00:47:16] Ben: Yeah I'm using granola as well and I feel like I'm just digital hoarding. I don't go back and use them, but they're there in case I need them. And it's what's the purpose of transcribing them?

It's like an in case scenario, which seem really relevant, but it's probably something I could or should be picking out of some of that. And I might need it. I might not. But yeah, I think it's interesting. Eric has a question and we'll finish on that. Oh, I need to meet you

[00:47:46] Guest: again. Thanks. I have a question about every business model. I was reading some article that you had written. I think it's like your growth objectives and you had this pyramid that was describing how you were intending to grow the [00:48:00] business and something that's stuck with me. That's been very interesting about that is.

Software is actually fairly low in the pyramid. So software courses, things that scale are lower in the pyramid. At the top of the pyramid is more advisory speaking type work that doesn't scale. And so I'm curious if you see Everee as more of an advisory firm or business model, or what conclusion I should draw about what

[00:48:26] Dan: you're doing from that perspective.

Great question. So the thing you're referring to is I wrote this article called every master plan and I have a pyramid. I think of every as a pyramid, hopefully not a pyramid scheme, but like pyramids shaped and at the bottom of the pyramid is like all the content that we do all the free content. Above that as our sort of like main offer, which is our every subscription is 20 bucks a month and you get access to our writing our discounts on our courses, our software, all that kind of stuff.

And above that, we have courses and training. And then above that, we have consulting. And [00:49:00] I think you shouldn't think of the top of the pyramid as our like focus. The top of the pyramid is the pyramid gets smaller as the percent as the number of people in the audience who are relevant. The number of customers gets smaller.

So at the bottom, we have hundreds of thousands of millions or millions of people who like, read or listen to our stuff. And then above that is we have a main offer where 20 bucks a month, 200 a year where you get all the stuff that we bundle together. And then above that are like, and maybe okay. Tens of thousands or hundreds of thousands of people will eventually subscribe to that. And then above that are like courses or trainings or advising or speaking where it's like, there might be only like five people a year that are going to pay us like a bunch of money for a training. But that's a lot of cashflow and we can take that cash and Push it down into the bottom of the funnel and expand to millions more people or whatever.

So that's more how you think of it. So rather than it being the like focus of the business, it's it's a function of number of people who are the target audience and the amount that they're willing to [00:50:00] pay.

[00:50:00] Ben: I know we're on time down, but there's just one question that I forgot to ask, which is from Diana talking about how do you approach the challenge of ensuring that AI has your personal taste without. But like without reforcing biases in sort of your creative work,

[00:50:16] Dan: that's a good question. Answer is it's impossible to like, not in some way be tilted or bias in a particular direction.

To some degree, and it depends on what kind of bias you're talking about, but in some, if we're just talking about bias in terms of having a particular perspective on the world, like to some degree, like that's important and you want it to push you in in perspectives that like feel helpful to you, obviously like bias more generally like actual bias is like not good.

But I think generally these tools will do what you ask them to do. And so if, Being biased in a particular direction is something you're worried about. You can ask it what's another perspective? And it's going to be, like, really good at telling you what the other [00:51:00] perspective are, or who the people that disagree are, or whatever.

And in general, I think it's going to be far less biased in that way than almost anyone you could talk to. So if that's something you're worried about, I would say just use it to break down your biases, because it will be good at that.

[00:51:17] Ben: Awesome. Thanks, Dan. I know we've taken just slightly more of your time than we planned, but yeah, I really appreciate it.

And it was great. I'll get this edited and up as soon as possible. But yeah, really appreciate you taking the time.

[00:51:31] Dan: Of course, really appreciate you. I love what you're doing. I love this community. Feel free to reach out if anyone has any questions and I hope I hope to be back sometime.

[00:51:38] Ben: Yeah, definitely.

Sounds good. Cheers, Dan. Cheers, everyone. See ya.

‍