Transcript of Joshua Wöhle: Turn AI Into Your Competitive Advantage and 50X Your Productivity | Artificial Intelligence | E393

AI is not going to take your job. Somebody else using AI will. If you walk into a job and you say, I would like a job, but I don't use AI, I don't think there will be a job.

Joshua Vohla is the CEO of Mindstone, an AI-powered learning platform helping professionals actually use AI to do better work faster.

From Copilot to ChatGPT to Claude, number one problem is people use them as search engine replacements. Talk to it like you would talk to a person. Ask it to ask you questions. Don't just ask it for answers. The number one role of a great manager is to make themselves obsolete. And actually, the same is true here. Yeah, you are building your AIs. Your job becomes building a system around you that can do the job you're currently doing without you so that you can go and do the next thing.

How should knowledge workers think about the fact that they might be training AI to take their jobs and now like they're suddenly judged on how well they use AI? Like how should they be navigating all that?

I think the biggest thing that people can do to prepare is to—

Is there anything in 2026 right now that you absolutely will not delegate to AI?

The reality of where things are at is moving much faster than how people are able to catch up. Hopefully, I mean, this is an interesting one because—

Young and Profiteers, I recently went to an AI retreat and I literally built an app in just 15 minutes and I have zero coding experience. I never thought anything like this could be possible, and my mind has not been the same since. I've been obsessed with AI and learning AI as fast as I can because now I truly understand how it's gonna change everything for us, how AI is changing the workforce, why most people are using tools like Claude and ChatGPT completely incorrectly, how to build systems that do hours of work in just minutes, and what skills you need right now to stay relevant to keep your jobs or your companies alive. This is not a theoretical episode today. This is stuff that you can actually use and apply today. In my opinion, this is mandatory listening for anybody who wants to be successful this year and for the years to come, because AI has taken over and you need to learn how to use it effectively. Without further delay, Joshua, welcome to Young and Profiting Podcast.

Thank you very much.

I'm so excited for this conversation. You know, I, I went to your Mindstoned AI Breakthrough weekend recently and my mind was absolutely blown from from everything I've learned. And so I thought I have to bring Joshua on to just unpack everything with AI, 'cause you just know so much and I'm just so excited to pick your brain today. So, so thank you so much for joining us. So one of the things that, um, you've said in the past is that AI is more dangerous from people who are ignoring AI than just from AI itself. The danger is in actually ignoring it. So let's paint the picture. If somebody is a, let's say a 30-year-old knowledge worker, a marketer, or even an entrepreneur, What does that look like if they ignore AI for the next 3 to 5 years?

First off, 3 to 5 years in AI feels like a lifetime at the speed that it's currently going. But the, there's this saying like AI is not going to take your job, somebody else using AI will. And I do think that is true, which is that Today, if you try to walk into a knowledge work role, any role in a company, and you tell them, "I'd love to get a job, but I don't use email," they're probably going to have a problem. And I think the same thing is going to be true with AI, which is if in, actually, probably in 2 years, if you walk into a job and you say, "I would like a job, but I don't use AI," I don't think there will be a job.

When I went to this training weekend with you, it was just a few weeks back. I felt like there was no way I could ever create an app myself, for example. I ended up creating one in like 15 minutes after some training with you. I did not fully realize the capabilities that AI had. Like I knew it was there, but like I thought it wasn't for me. And, um, of course I used ChatGPT and all the basic stuff, right? But like I didn't realize like how far ahead it already was and how far behind me and my company were at the time. So talk to us about what was the mindset shift for you when you were like, okay, AI is going to change everything. You know, what capability did you see it do where you're like, okay, this is really going to change everything and we need to, you know, hop on this trend right now?

So I think there are a few things. One is I am a technologist by nature, I'd say. So I guess I kind of fell into it a little much earlier than most. I tend to want to figure out how to make it work. So that was the first thing, which that is different to most, and that explains why maybe we were there earlier. The first time that it really clicked that I thought, okay, this is going to change everything was when with Mindstone, the original vision was about general learning. We wanted to help people learn in a way that would be more effective, uh, and that really would move the needle, that helps them have better lives. And at some point, and it was, this was in between GPT-3.5, which was the original ChatGPT, and GPT-4 when it came out in March the year after, um, we realized that AI could estimate the approximate level of understanding of a subject for any individual based on the quality of the question they were asking. And that was fairly meta because everyone was using ChatGPT to ask questions and sure, you get answers back, but the fact that it could derive your level of understanding based on the type of questions you were asking was a level behind.

And suddenly it made me think, what are some other things that are completely non-obvious that you might want to ask this? AI, and that was really the starting point. I mean, it's been basically 3 years of discovery after discovery since.

I wanna dig deeper on that because at the AI retreat, I remember one of the exercises that we did was basically, um, asking, uh, your AI of choice, um, tell, ask me one question at a time to get as much context as possible before answering this question. And to me that was like genius. And now every time I ask ChatGPT something, before I even ask it the question, I first say like, ask me a question, you know, one by one to fully unpack context before you give me an answer. And this was really powerful because before I used to get really frustrated with AI and I would like throw away a lot of the answers it gave me 'cause I'm like, oh, it just doesn't understand. It doesn't have the full context. So talk to us about like how we can prompt better in that way and why that's so important.

Yeah, so that is one of my favorite techniques as well, is the asking the question bit. And a lot of this actually comes back to changing the way that we interact with AI. Now, one of the frameworks we teach a lot is stop thinking about AI as technology, think about it as a person, which is both right and wrong. I'm not trying to say here that we have some kind of being there, but from a framework perspective in how do you interact with the technology, it actually works against us that we have a past of Google and how you ask questions in that way. When you have a conversation with a person and you ask that person a question or you want the help of a person, you naturally give that context. If we sit together and I asked you advice on how to run my business, if you genuinely tried to help me, one, you'd ask me, 10 questions first to get that context. Now, the AI doesn't do that automatically, which it should. And honestly, sometimes I wonder why the hell did they not build it in a way that it would do that naturally?

But anyways, it doesn't, which means you have to create that context. And you have to create that context that mimics the way that you would have this conversation with a person. After the context is set, the quality of the answer that comes back is going to be much higher. Because the opposite is also impossible. You cannot get a perfect answer if you do not give the context. Literally the best person in the world wouldn't be able to answer the question, how do I, or what is the better strategy between two different strategies for my company if I'm not giving them the context? It's just an impossible situation. But because we are used to using technology in that way, people judge it in that way. So if you ask, okay, which one of these two strategies is best and it comes back with a bad answer, the default is people think, okay, well, it just can't help me. Which is, and this is the problem that you then come into, which when you have a person conversation, if the first answer you would give me isn't quite what I was looking for, my answer wouldn't be, okay, you're not worth having a conversation.

I would say, no, actually that's not quite on point. And this is why I think you're missing this piece of context. And I would explain it to you. And then we'd continue the conversation.

Yeah. So moral of the story, yap bam, is that when you want to get ChatGPT or whatever AI to help you, you actually want to ask it, please give me questions to understand more context before you actually, uh, answer the question, right?

100%.

The other, uh, mental shift that I had was actually talking to AI at the retreat, right? So I was so used to chatting all the time, but you guys actually directed us to, uh, they were called walk and talks and basically just talk to AI and that really helped and made it feel like it was a partner, a real, like, you know, teammate helping me. Why is that such a powerful unlock?

So this goes back to, I guess, one of the advantages of having started out as a general learning company. There are just different ways that your brain interacts and that your brain acts when you're thinking, when you're interacting with ideas, different neurons that will fire in different ways depending on the context that you're in. That context can be the room that you're in. It can be the method that you're in when you're walking or when you're sitting. It can be the method that you're using to interact with it, talk to it rather than typing on a keyboard. And so what happens when you talk to AI and what we did on the weekend is literally just put your AirPods in, go and have a walk around and talk to the AI. In a way that you would to a person, all your surroundings are now different. And so the ideas that come to you, the way, the things that you will say are going to be different than if you just sit behind your laptop and type it into a chat interface. And what is important to recognize there is that actually both are important.

It's not necessarily just walking around is the only thing you should be doing. But it is that you see different sides, and so you want to combine them. And that's where the magic really happens because suddenly you get the rigor of when you're sitting and kind of the work that you do on a daily basis, but the creativity of walking around and the ideas that otherwise wouldn't have flown. And then the AI is able to take both of those into account and give you much better recommendations.

I also feel like the AI is designed to be more succinct and conversational with the voice. Prompts. And so it's also like a quicker kind of getting to an ideation, to your point. Like the first thing I thought of that's different is like, wow, this feels so much more creative and like natural.

Yeah, absolutely.

So how has AI changed the way that you work as a CEO?

You can almost frame that the other way around, which is I don't think there's a single thing that I don't do differently.

Yeah, I can imagine.

So the— I mean, the biggest parts are I no longer make a decision without running it through AI first. It just feels like trying to do complicated arithmetic and not using a calculator. Like, even if I knew how to do the math, I would still use a calculator to make sure that I'm not making mistakes, right? That would be the biggest thing. The second biggest thing is being able to take into account all the context of everything all the time. So the biggest strategic partnerships we talk about, the biggest deals we have with customers, investors, the ability to draw on every single conversation, every transcript that I have of my conversations with them, and then being able to surface what are the most important aspects based on any relationship that we have built with customers, investors, or anything else. That is just something I wasn't able to do before. I can't hold 10-hour-long conversations in my head at any point. Maybe I can remember some of the important points of the last 2 meetings if I'm lucky, but my memory doesn't stretch far enough. And so I would lose a lot of detail. Now I don't.

I want to explain to my listeners just how life-changing it was to be at this retreat and to use Rebell, which is your tool at Mindstone. So before I get into it, talk to us about like what Rebell is.

Many people will have tried ChatGPT and Claude and all of the really cool tools that exist. That would be kind of a chat agent and it works really well, but it's somewhat superficial. It doesn't really know who you are. It doesn't really know what your company does. And so first and foremost, Rebol is just a really, really evolved agent that is able to do real work rather than stay within its own confines. It's able to connect to all of your systems. It knows you inside and out. And then second, it's an AI-first operating system for the company that you are in. And so it is building the infrastructure for the type of organization that you need once every person in the company has their own agent. And those agents need to collaborate, and they have shared memory, so it all compounds. You need the safety mechanisms, you need the privacy mechanisms at the same time, because I wonder if the— if people are familiar with OpenClaw or Claude Co-Work, both of these are starting to come up. They have their real privacy issues because they do a lot. But once you start to put this in a company context, people need to know that the system is able to do the work without having too much of the collateral damage.

And so that is what, what Rebol really has been built to be, is the infrastructure that allows every person in the company to have a capable and trustworthy agent whilst at the same time, helping them collaborate, having shared memory and so on.

Young improfitters, raise your hand if you've ever Googled one tiny symptom and then 5 minutes later, you're convinced that you're dying. I know I have been there, done that way too many times, but instead of guessing and spiraling, I use Zocdoc. In fact, I've been using Zocdoc for like 10 years. I literally don't know how to book doctor's appointments any other way. It's a free app and a website that helps you find and book high-quality in-network doctors based on your needs. You can search by symptom or specialty, read verified patient reviews, and see doctors with available appointments right away. If you're a busy entrepreneur and you need to squeeze in a doctor appointment on a random time, you can find the doctors that have availability when you have availability. Want an in-person or video visit? It's no problem. With Zocdoc, there are over 150,000 providers across all 50 states and over 200 specialties. Any doctor you could imagine, any visit that you need, Zocdoc will have you covered. And most appointments happen within 24 to 72 hours. So next time you start doom scrolling your symptoms, skip the stress and book a real appointment with a real doctor on Zocdoc.

Stop putting off those doctor's appointments and go to Zocdoc.com/profiting to find and instantly book a doctor that you love today. That's Zocdoc.com/profiting, Zocdoc.com/profiting. And thanks to Zocdoc for sponsoring this message. Hey, YAP fam. I'm not afraid to say producing this podcast requires skills I do not naturally have. From audio engineering to video editing, I have to hire experts who are way better than me in those areas. And as I scale my podcast, this becomes more and more important. And every time I need to hire, my first thought usually is, this is a job for Indeed Sponsored Jobs. When you sponsor your job on Indeed, you find candidates with the exact skills you're looking for without the stress of digging through endless resumes. Sponsored Jobs posted directly on Indeed are 95% more likely to report a hire than non-sponsored jobs. That means you're not tossing your post into the void, you are connecting with qualified people who can actually help your business grow. Spend less time searching and more time actually interviewing candidates who check all your boxes. Less stress, less time, more results. When you need the right person to cut through the chaos, this is a job for Indeed Sponsored Jobs.

And listeners of this show will get a $75 sponsored job credit to help your job get the premium status it deserves at indeed.com/podcast. Just go to indeed.com/podcast right now and support our show by saying you heard about Indeed on this podcast. Indeed.com/podcast. Terms and conditions apply. Need to hire? This is a job for Indeed sponsored jobs. Yap gang, it's confession time. I thought I had my money handled. I thought I had a good handle on everything going on, but I recently checked my bank statements and there was subscriptions everywhere— streaming, apps, random charges. A few of them were billing me for a long time. I didn't even know what they were, and it was a painful uncovering. So I downloaded the Experian app, and now it's like I have a financial assistant in my pocket. The Experian app helps you track spending and find subscriptions that you forgot about and you no longer want. With your subscription cancelation feature available with an Experian Premium membership, you securely link your accounts They scan for recurring charges and you pick what goes and what you want canceled, and then they cancel it. It's as simple as that.

They have over 200 eligible subscription types, and it's usually done in 3 to 7 business days. If getting your money tight is part of your growth plan this year and you don't wanna manually have to do it, this is your move, young and profiteers. Get started with the Experian app today. Results will vary. Not all bills or subscriptions are eligible. Savings not guaranteed. Paid membership with connected account required. See Experian.com for details. Okay, so just so I can walk you guys through like how I've used Mindstone since 2 weeks ago when I first found out about Rebel, basically Rebel's a platform, Mindstone is the training company essentially. So Rebel ingests all of my emails from the past and like also ongoingly it takes all of my G Drive files and, and basically ingest them as well. And it like stores it as memory. I hook up my Fireflies transcripts. Um, I uploaded all the old ones and then it ongoingly processes the new ones. It can scrape all of my social media and get all my post content. And basically it, it can, it can sync up to Airtable, it syncs up to my finance hub.

And essentially it has all the memory of like how I talk in emails, how I like all the conversations I've had, every contract that we have. And now I have essentially this second brain and I have it do automations where it essentially flags anything that I have to do in my inbox and it will, it will check my email twice a day. My Slack, by the way, it connects to Slack and all your messages and basically I can say a voice note and be like, respond this way. And then it will send a draft in my email that I can just tweak and edit and send out. So now responding to 20 emails takes 5 minutes instead of 20 hours or whatever, like, you know, 2 hours or whatever it would've taken. I don't miss my Slack messages anymore. It's making me a better leader. It's totally transformed how I operate and I'm so excited to like keep on using it. And so much so that I'm literally taking every single team member Even the, you know, everybody on my team's super important, but even the most junior person on my team is gonna take the Mindstone training and I'm gonna be implementing, you know, a hiring freeze because there's absolutely no, for me it's like, okay, what are we gonna do?

Because I've just unlocked so much capacity. Um, and so one of the things, one of the rules that I'm thinking about implementing is before we put out a JD for a new hire, we've gotta prove that AI can't do it first.

Ironically, you might want to ask Rebel every single time, can you do this? But yeah, absolutely. It's, it is, we are in a very weird period at the moment. I love the fact that we're able to do these types of weekends because there is, this conversation needs to happen much more in the open at the moment. Right now, the public discourse is still The majority of world thinks that it's somewhat of a hype, that it's maybe, it, it can't possibly be all that people are making it up to be.

I mean, I thought that and I interviewed the biggest AI experts in the world and not that I thought it was hype, but I just didn't think it was actually like executable now for me.

Yeah. And, and, and that is exactly the, and that's why the, the weekend works. Uh, and I was talking to a few friends about this before as well. The hard thing is the claims of the technology are so big that it gets instantly ignored. Like when I tell people that I genuinely can now do 50 times the amount of work that I was able to do 4 years ago, that just gets, I mean, it gets laughed out of the room. For those that, and the problem is it's very divisive. For those that know, they know it's possible. And so then you get in the conversation. For those that don't know it's possible yet, it's so outlandish that you're not even given the time. I can't show you in half an hour how you get there. But over a 3-day weekend, I can start to show how everyone understands it's actually possible. And that is why those weekends are so powerful. But that is also why it's so important to have more of this conversation now, because it means that 99%, if not more, of people are currently not aware of that is what is possible.

And so that means that the rest, the next few years are basically going to be about people that are able to bridge that gap and people that move slowly or don't move, don't bridge the gap at all.

Something else that I've been using Rebble to do is to essentially create these skills where I ask, I say, hey Rebble, ask me a question one by one to unpack how I do X process. And then I say, okay, now that you know the process, you do it, right? So talk to us about the concept of skills and agents and how, especially for somebody who's never heard of any of the— how should we think of this? And how can somebody get started creating their agent?

This is actually a really interesting bit because a lot of companies exist today that would have what you would call vertical agents. You've got a customer support agent, you've got a finance agent, whatever agent, and you'll have thousands of them in different ways. Actually, a skill in Rebol is not too dissimilar from what people would call agents, with specific qualities and examples and capabilities. And so, we believe fundamentally we will likely not end up in a world where you'll have a thousand different vertical agents. You end up with one agent that has very thoroughly developed skills, and it can decide to use those skills as and when it sees fit. So it has a skill to manage a financial Excel sheet. It has a skill to help build great pitches. It has a skill to build proposals. It has a skill to build presentations, whatever the thing is. So a skill is a combination of a process, basically a process that the AI can execute reliably. You tell it, do X, Y, and Z every time I tell you to execute this skill. It is a set of examples so that the AI understands what does good look like when I am asking for a proposal or whatever the thing is.

And then sometimes, and this gets a little bit more advanced, sometimes a skill can include some scripts, which is basically some deterministic steps, some things that always, that it can double-check for itself to make sure that it can consistently execute on the skill as you go through. So for Excel, this is one of the things because it's managing an Excel sheet is a very complex set of actions and sometimes it has some kind of underlying programs, if you want, that help the skill execute exactly what it's supposed to.

I found this skill so helpful. I created like a lead prospecting skill And it, it's, Rebel's able to like scrape emails. It can, uh, I say all the things like the qualifications, like, uh, for me I'm trying to recruit podcasters. So how many downloads do they get? How many reviews do they have? Links to their social. Then it will even go as far as crafting a custom email, grabbing the emails and then drafting them in my inbox. It is insanity. It's like, it's literally like having a chief of staff that actually has full context. Of your entire work-life history, as long as you were able to like provide that memory for it.

Yeah, so to your point, actually, it's a very good point. The skills can be executed when you want them to and also automated, right? To your point, the automations. And so one of the skills I think that I built live on the weekend, which I'm now using every day, was this, I think it was a lost deals evaluator skill. So this was, just one of the ones that I hadn't gotten around to building yet, but just something that looked at my Slacks, my emails, my WhatsApps, and looked at deals that were in our HubSpot that didn't have activity for a while, and then drafts an email to revive the deal that I had forgotten about or I had let go for too long. And I remember I did— there was one of the people that was on the weekend that was actually part of the emails that it wrote live, I think, whilst we were there, which was a fun bit. But that That is indeed a good example of a skill.

I mean, I've been using it so much to the point where I'm like, okay, this is supposed to save me time, but I'm actually working a lot more now because I just feel like, oh my gosh, this could be a skill and this, and, and I can be a better leader by, you know, using Rebel in this way. And so I found myself after the weekend, I think I got home on a Saturday, Sunday I was supposed to like have like a, a day off, but I found myself working the entire day on, on Rebel because it was just so exciting to me. So what kind of boundaries are you putting in place? Uh, like how are you ensuring that you're learning AI? Uh, and of course I know it's your, you know, main thing, but for like the normal person, how should they ensure that they build in time to learn AI and how can they also put boundaries if they find themselves like me staying up really late and working through the weekends because it's just very exciting.

Something that we're, we have recently become more aware of and we're spending more time on. Actually, the end of the weekend was exactly about that, right? Just where you have all this newfound ability, where do you actively want to deploy it? Because most companies don't actively think about that. They go down the productivity line and they get the extra time, but then they're not as conscious of where that time gets deployed. And that is important both at a company level and at a personal level. Now, there are a few things to look at here. And it's, I think there is a part of this is an individual answer for everyone, because not everyone will find the same balance. This whole thing about work-life balance is different for whoever you talk to. But the thing that is most important is that you consciously go in. And I think Jeremy, who was my co-host on the weekend, he puts this fairly well, which is thinking about the person you want to be and being clear about what that person is, who that person is, and then using that as you get introduced to AI, as you start to find different optimizations, as you find more time back so that you're clear about where you want to deploy that time.

A very active bit of that right now is, I mean, I'm trying to live part of it at the moment, right? So I've got, we're here in Austin, I've got my wife and kid here as well. And for the first time ever, I'm basically trying to combine, he's got 3, 4 more months left before he's starting school. So this was the last time that we could go and get some travel in it and just trying to do it all really, which is, running the business, being in the US more, most of our clients are US, but also being able to travel a little bit going through. And so in 6 hours a day at the moment, I think I'm able to do what would have taken a full week before, almost forcing myself to live that future so that I'm not setting the bad example either, which is that because it is very easy to get drawn into this being 24/7 because it's very addictive once you're able to do so much stuff. I mean, we all want, ultimately we all want to be productive is maybe the wrong word. We all want to be good at what we're doing.

So when you suddenly feel that you can be twice as good at it, 3 times as good at it, 4 times as good at it, that feels really good. And it's actually very addictive.

It is addictive. I feel like we need to make sure that we're building the right things and not just building to build. The other way that I've been thinking about it is like, so when I first, I've been an entrepreneur now for like 6 years. The first 4 years I worked myself to the bone. I'd stay up till 2:00 AM every morning. I worked through every weekend. And then suddenly like I built all this capacity and a team and trained everybody where now I can do whatever I want. You know what I mean? Like, you know, I still work hard, but I can take a week off. I can do it. I, I've, I built a business that runs without me. And so I think of it as this, like, you gotta like set everything up and now is a time to like focus and figure it out and think about like, what can I automate? How can I redesign my team where like these low priority things that we used to spend a lot of time on, now we use AI to do them, and how can I open up my team for more capacity?

And it kind of feels this like this new like reset of I need to set my team up for success and this might take some focus time now, but I imagine in a year after I've, I've set a lot of these things up that it would free up capacity. So it's kind of like this like sacrifice work hard now so that you can free up your time later on. Is that the right way to think about it?

Yeah, for those that want to create that opportunity for themselves, absolutely. I mean, that's definitely what I am doing as well. So the putting in place that basis, and it was, I think for me, there was a tipping point somewhere. When was it? I think it was October of last year. Up until that point, I think I probably I have never worked more in my life. And so I was doing whatever, like 18, 19 hours a day, every day, 7 days a week because I was in that loop. And so I was like, okay, I'm twice as productive, 3 times as productive, 4 times. I'm 10 times as productive.

Shit.

Like, and literally was going through—

You're like, how far can I go?

Yeah, exactly. And then I think there was, yeah, so in October, there was a very conscious moment where I realized for the first time, I was like, Actually, I'm gonna, next week, I'm going to consciously start to do less. So I felt, I mean, I'm in a different place as well because I've, I'm on that cutting edge. So I kind of felt that it's okay to only be 10 times more productive than what I was before. Nobody else is there yet, so I can take a bit of a breather in a way.

Yeah, help us all along get to where you're at. You're a very successful entrepreneur. You've had multiple companies. So you had Super Awesome that you sold to Epic Games. Um, and at that point you could have just retired a little bit or taken a break, but you jumped straight into starting Mindstone. So what was the genesis of starting Mindstone and talk to us about, uh, why you pivoted into AI?

My view of a great life is not one of sitting on the beach and not doing much. Um, that's just not who I am. And so the biggest thing is I knew I wanted to do something that would make a difference. Actually, literally the number one way that I think about measuring my life is leaving a legacy in one way or another. And then I hesitated between a few industries and the one that I, I think it took about two months or so of kind of thinking through, the one that I thought I could really make a difference in was education. Because I mean, it is entirely broken. There's so many things that need to be different or even already needed to be different six years ago. About education itself. And I have a bit of a weird past with education. I always loved learning. I hated education, thought there was a better way to go and do that. And so really, the first bet was just, okay, I can see myself spending 30 years tackling problems in education and try and be as helpful as I possibly can. And then about 3 years in, well, ChatGPT happened.

And it also happens to be actually just before starting SuperAwesome. And this is I think this is actually something very few people know about me yet, but I was supposed to go and do a master's in AI in London. I'd been accepted there to go and do the thing. And that was going to start in September. And then I think we had conversations with Dylan in April, May of that same year about starting Super Awesome. And so ultimately I decided to go and start Super Awesome, but I've always had this inkling of AI. I've always been really passionate about AI to begin with. And so somehow when then ChatGPT happened and suddenly everything in learning changed, both in terms of what we could build, but also in terms of the skills that would be useful in the future, it's kind of like every single thing that I really like doing in life came together.

That's amazing. And it's a good thing you didn't take that AI program because probably would be completely irrelevant today. But to that point, What are the things about AI that you learn that continue to be relevant? Like, so for example, you've got this Mindstone AI training.

Yeah.

I assume if people take this training, it's not gonna be irrelevant in a year because there are foundational things that you need to learn in terms of how to use AI. So talk to us about some of those things.

We've touched on a few of them, which is talk to it like you would talk to a person. I think that's the, the number one. The second one is ask it to ask you questions. Don't just ask it for answers. It's not terrible to ask it for answers, but actually the majority of your interactions should be getting it to ask you questions instead. Help it help you think. Don't let it take the thinking away. If you're a manager, there are things you can apply of how great managers work that you can also apply to AI with a slight edge. So one of my other favorite ways of using AI is getting it to rate itself. When you're building companies, you're building teams, if you're successful at what you're doing, hopefully you're hiring people that are better at their respective jobs than you are, right? That's why you're hiring them to do those jobs. So it's really hard sometimes to then give them great feedback because actually they know their job better than you do. And so one of the main techniques of a manager ends up being, person comes back with a piece of work, it's great.

How would you improve on this particular piece of work? They go and line out all the things that could be better on that particular bit. Okay, great. I agree with you. Now go and do all the things that you just said you could do to make this even better than what it already is. In a human context, you have to think about how does that land with the person? They just spent a week doing this piece of work. I'm now asking them to spend another week. How are they going to, enjoy this, they have a whole bunch of other things that need to get done. So you have to think about that emotional component. With AI, you do not. But you can apply the same technique, which is ask the AI, okay, you just did this piece of work for me. On a scale from 0 to 100, how would you rate that you just did that job? It comes back, I think I did a 78 out of 100. It's great. What is wrong with it? How would you get it to 100 out of 100? Well, I would do all these things. Okay, now go and do it.

It really is like managing someone.

Exactly.

Yeah.

That's another really big one. And I think those few ones are already things that everyone today can apply because they're simple, but extremely powerful. And then once you start to get into the habit of using it in that way, you're then starting to get to the edges of agentic AI and like getting it You've got these custom GPTs or projects on Claude, or skills are actually not too dissimilar from a custom GPT or a Claude project, but just the agent decides to execute them when they can. And there, I'd say this is the encapsulation of certain units of work. So once you start using it more regularly, then you have to start thinking a little bit more inward, which is what are the things that I am doing that I'm doing fairly frequently and that take a decent amount of work, or the things I should be doing frequently that I never get to or that I always do to a low standard when they should be a high standard. Actually, that second bit is the thing that people often forget because it's a treasure trove of potential AI enhancement. It's like, what are all the things that if I had the time to do, I could go do and that would make a big difference on my job?

Or on the quality of my work? Well, get AI to do them. If you don't have the time to do it, get AI to do them well.

It's so interesting that you bring that up. I remember one of the questions that you guys asked on the AI weekend is that you, you, you say AI is like an intern that doesn't sleep, right? And you're like, if you had to give an intern something to do for 24 hours a day or even 12 hours a day, would you even know what to give it? So how can we start to understand like what we actually need to build?

Yeah, this is where people that have managed people before do have an advantage, uh, and people that haven't quickly have to figure out what that means because the, we are all basically becoming managers in one way or another. Uh, and I would also argue that the analogy of the intern is very quickly becoming the mid-level employee and the senior employee.

Yeah, I agree.

So, depending on the quality of your Rebel installation, I am now, like, when I look at our proposal skill, it does better proposals than I do, no question. And I would consider myself a very good salesperson in one way or another. I do good proposals. It's a skill that builds up over time. And this is the Both the good part and the bad part, which is that it's actually very easy to get going. Luckily enough, the technology is not the problem. We often talk to customers when we start working with the organization, we say 5 to 10% of what we do is helping people use the tools. 90 to 95% of what we end up doing is helping people with the mindset shift of thinking differently about the way that they approach work to begin with. And that is where you get the real return on generative AI. And it's actually the thing that people forget. They think that they can just roll out ChatGPT and people will just figure it out. It's as if you think, okay, well, we're going to give every, to your point, we're going to give everyone an intern and they'll just figure out how to make that intern really useful, let alone if you're giving them actually capable, really capable employees, like they wouldn't know what to ask them to do to begin with.

And so the real work is helping people understand how they can take advantage of that technology.

There's so many ways I can take this, but I think the next point that I want to talk about is the ability for the workforce to be able to code now, essentially, and build apps. So I was totally blown away by the fact that on this AI Breakthrough weekend with Mindstone, I was able to basically create an app in 15 minutes. It basically took a podcast episode, I asked, you know, it was ChatGPT, I asked it to uncover my, my process and to, to help me come up with a prompt that I then put in Ripplet, which you guys taught me about to create the app. And then now I have something where I can pop in my episode transcript, link to YouTube, it pulls out the hooks, the clips that we should use, the best quotes, like all this kind of stuff. I also created a new app right after the weekend, of course, to help me create research briefs. And honestly, the output of that app is so good. So much better than what my team was giving me, what I was doing with ChatGPT. Like, it is just so much better. Um, and I was talking to— I met this girl who works at one of my sponsors, actually their company, so I'm not gonna say the name, but it's a very big tech company.

And she's in sales, and this company is very AI forward. And she told me that they have an AI quota. Okay? They actually get judged by how much AI they use. She's a Salesperson, she said she's coding now and that, um, they actually will fire people who, if they find out that they're doing stuff that AI could have done. Yep. And it's, it's basically, she's like, uh, she feels actually scared because she's, she's really excited that she's learning all this AI. She knows that if she, if for some reason she loses her job, she's so much— when I was talking, I was like, damn, you're really like they're preparing you for the future. So either way, it's a good thing. Yeah. But she's also scared because she said that, uh, part of it is like she gets AI-generated emails, like in her inbox before she responds to an email, the email's already drafted.

Yeah.

And then she actually edits it before she sends it out. So she's actually training the AI and she basically said, I'm kind of training the AI to take my job. Yep. So there's a couple questions I have in, in this whole breakdown is like, How do you imagine the workforce in terms of their ability to code? What does that mean for software engineers now? Like, what is the software engineer role turn into in the future? Yeah. And then also like, how should knowledge workers think about the fact that they might be training AI to take their jobs and now like they're suddenly judged on how well they use AI? Like, how should they be navigating all of this?

Yeah, this is where the real questions, yeah, start, right? So the, The interesting part is we're coming back to the analogy of what does a great manager do? It's been the case for years and years that the number one role of a great manager is to make themselves obsolete, is to build a team that can function without them. And actually the same is true here. Yeah, you are building your AIs. Your job becomes building a system around you that can do the job you're currently doing without you so that you can go and do the next thing. And we'll get through to what that next thing then becomes soon. But when you think back from a software engineering perspective to the first question you had there, I think it's really useful for people to look at what is happening with software engineering. And it's hard because people have not emotionally been, most people in the world are not emotionally connected to software engineering in one way or another. It's actually been a, like, software engineers had a really good life, were getting paid a lot of money to go and do things that they actually loved doing, but in some ways they're absolutely living the future.

And when I say that, if you are still manually coding something by hand today, as a software engineer, you are already behind. Nothing we do in Mindstone is manually coded anymore. Actually, we're now at the point where almost everything doesn't even get seen by a human anymore from a code perspective. And that seems outlandish, um, but it is definitely what is happening everywhere. And so as a software engineer, your job used to be coding, say 10 years ago. Now, having said that, Last year or 2 years ago, already a great software engineer, their job wasn't just to code. You can maybe think 30-40% of their job was to code and the 60 or 70% was to figure out what to code to begin with. So a lot of even software engineers then took what is happening at the moment in somewhat of a bad way, or at least they weren't extrapolating where they were looking at, okay, the coding part of my job is now getting automated. Great. I can do other things, which is exactly the attitude you should do. So, hey, do more of the other bits. But what is happening this year is also the bit of figuring out what to build is now also getting automated.

Figuring out how to build the spec, figuring out the architecture, all of these pieces that were higher-level software engineering that a senior software engineer or an architect would do, they're also now getting automated. And where this is so useful as a paradigm for everyone else, is to think through what is the equivalent of coding for a software engineer, but for someone in finance or for a lawyer or for an HR professional or a marketer. The equivalent of coding for someone in finance might be managing an Excel sheet. That's disappearing this year. Probably by the end of this year, maybe even the middle of this year, you shouldn't be manually updating your Excel sheets. That means that you can live one level higher. Um, and that one level higher, the, the strategic financial picture and everything, um, maybe the supply, the way that you deal with suppliers, the way that you look at the financing structure of the business, that will still have a year or two, uh, extra on top before that also starts to get touched in different ways. So taking that bit really seriously, I think, is probably the most important bit because people keep discounting this as if it is fairy talk.

It is not. My CTO, Greg, recently did, it was literally 2 days ago, did a talk at a CTO conference. It was very interesting because this is a really forward-leaning audience. You would expect them to really know their stuff. And so he had prepared this talk thinking, okay, I'm gonna go into a whole bunch of people that are very on top of their stuff here. He had prepared the starting point of his presentation to figure out where is the room at from an AI adoption perspective. The first question he asked was, how many of you are still writing code by hand all the time? So without AI. And he had expected, okay, that's gonna be very few. Most of the room had their hand up.

Oh God.

And he was like, Okay, how many of you are using at least agents to do part of what you're doing? Like barely any hands. And he had like 3, 4 levels behind that, which was how many are still, how many of you have an automated process to deploy what the agent has built? How many of you are still looking at the code? How many of you are still looking at the reviews of the code? Yeah, but he couldn't even ask them because even in that room, people are putting their head in the sand. It is the reality of where things are at is moving much faster than how people are able to catch up. And I think the biggest thing that people can do to prepare is to take note of what is happening there in software engineering and figure out what that means for their own job. Don't wait, at least for a little bit of time, live in a world where you're thinking what if, sure, make up your own mind based on what you think is real and what is not. But the number one thing is try on a regular basis to check back in if the technology can do a thing.

And also if you do it once and it doesn't quite give you the thing and you're like, oh, it can't really help me in this thing, it was kind of there but not quite what I was looking for, Try it again in a month.

Yeah, things change so fast.

Yeah, and I think that is the really important bit is getting people to understand this. So figure out what is the equivalent of coding for your own job, and then start to figure out how AI can do more and more of that. Try to figure out where you can then spend the additional time. And this is where we come back to you being deliberate in where you want to spend that time and where it is most effective. Don't let that happen to you. Be an active participant in that, in that change.

Yap Gang, on the show I've sat down with Robert Greene, Seth Godin, Alex Hormozi, Gretchen Rubin, James Clear, so many epic, epic guests. And what do all these guests have in common? Every single one of them has a bestselling book, and every one of those books are on Blinkist. Here's something I've been thinking about with AI. Everybody has access to the same answers. But the people who actually read, who absorb real ideas and think for themselves, they're the ones with the original take in the room. AI can summarize, but it cannot replace the frameworks that you build by reading widely. That's your real edge as an entrepreneur. Blinkist turns the world's best nonfiction into 15-minute reads or listens. Over 9,000 titles covering entrepreneurship, marketing, psychology, habits, all the things we talk about on this show. I use Blinkist to go through several books a week, pull out the key insights, and then I decide which ones I wanna read fully between meetings at the gym. Whenever I have 15 minutes, I'm listening to Blinkist. In a world where everybody is copying the same AI output, reading is the thing that can make you different. Grab your free trial plus an exclusive 30% discount at blinkist.com/profiting. That's blinkist.com/profiting. Yap gang, here's a reality about scaling a business that nobody's talking about.

More vendors, more invoices, more payments flying around. And if you don't have a clean system for managing bills, things can get messy really quickly. And that's where Intuit QuickBooks Bill Pay comes in. It helps you manage and pay your business bills directly inside QuickBooks so everything stays organized in one place. You can see what's due, control approvals, and understand how payments affect your cash flow. Vendors can also add their payment details, which saves you from having to chase them down for info. Instead of spending hours managing payments, wouldn't you rather be focused on building and scaling your business? Start paying bills the smart way, not the hard way. Learn more at QuickBooks.com/BillPay. Again, that's QuickBooks.com/BillPay. Terms apply. Money movement services are provided by Intuit Payments Incorporated, licensed as a money transmitter by the New York State Department of Financial Services. What's up, young and profiteers? When I started building YAP Media, I had to figure out everything on my own. Branding, operations, finance, marketing. Some days it felt like I had 5 different jobs and a never-ending to-do list. Now, as the founder and CEO of an almost 8-figure company, I can tell you that running a business can be extremely overwhelming.

What makes the biggest difference though, is having the right platform from day one. And for millions of entrepreneurs, that platform is Shopify. Shopify is the all-in-one commerce platform that helps you build, run, and grow your business. You can design a beautiful online store with ready-to-use templates, use AI tools to write product descriptions and marketing copy. You can use it to run email campaigns, manage inventory, handle shipping and returns, and so much more. Shopify lets you handle it all. Everything lives in one dashboard, and that means less chaos and more time actually growing your business. Start your business today with the industry's best business partner, Shopify, and start hearing— sign up for your $1 per month trial today at shopify.com/profiting. Go to shopify.com/profiting. That's shopify.com/profiting. As we're thinking about careers and navigating careers, some of the things that I'm kind of absorbing from this is if you can try to be in a company that's actually very AI forward, like that's gonna be great for your future, even if it feels painful in the moment, like being forced to use AI. I know you do something similar at your company, we can talk about that, but having a company that's very AI forward is actually a really good thing for your career.

Even if it means you get laid off in the future, you're gonna be way ahead from the other people who weren't put in that position to learn AI. So it's like, And then also, if that's not possible, I would say go find a company where that's possible. If that's not possible, learn it on your own. Do the steps that you need. You have a 4-week training that anybody can sign up for. It doesn't have to be through a company. So learn AI on your own if you have to. And then lastly, it feels like entrepreneurship and being a founder is the safest route.

It is. There was literally just as I was coming here today, there was an X interaction where someone was asking, if everyone becomes a founder, who are all the employees? And my answer to it was, well, agents. Agents are the new employees. But I wanted to focus on one bit that you talked about, which was this, the fear around the adoption piece and like making myself obsolete in one way or another. I think that is really important because there are multiple aspects in this. And there are, um, I fully agree with you that if you are in a job right now and a company that is not fully embracing this, I would say that is the number one warning sign. It's like, if you can get out of that job—

That company's going to go out of business.

Yes.

Even if they're not going to lay you off, that company's not going to—

Well, and they're going to let you get— basically, you're not going to have the job because the company is going to be out of business.

Exactly.

And you will not be in a position to easily find a job anymore because you've lost the skills.

And you're behind.

Yeah.

Yeah.

So that is, that is a really important piece, I think. Then within that, one of the things that I, I must say I struggled with is that when we started Mindstone, I thought our mission, our entire purpose was to light a fire in every single person in the world and somehow help them figure out how they could enjoy learning, love learning. I love learning. And I thought for a while that was what we needed to do. Because if you love learning, everything becomes much easier. I realized about halfway through, a little bit less, I think it was first 2 years or so, That was actually a fairly, or a very self-centered way of going through life. Somehow I thought that was something that everyone would like. But actually, learning is like many other things in life, which is that some people love it, some people hate it. And I realized, and it was so, it was a real epiphany for me, when I realized that actually most people are scared of learning, are scared of change. It became so obvious, and I, like, some people must be listening to this as like, well, obviously. So this was not obvious to me.

Um, when I realized that most people just don't enjoy that process, it really changed the way that I thought about learning to begin with. And then when AI hit, it became obvious that a lot of company leaders are, I would say, actually negligent today. When you have a team that you are responsible for, you might think that you're doing them a favor by not being clear or not giving them what you would consider the extra load of asking them to learn how to use AI in their job. You might think, okay, I can't, they're already stretched, they're telling me they don't want to. Actually, the kind thing is to tell them to do it. They will thank you for it afterwards. And a good leader is not just doing what their employees ask them to do, they are trying to get them to the best possible outcome. And there is a real dereliction of duty at the moment. There's a real failure of leadership, I think, in many organizations where they— because learning has been for the last 20 years, most companies look at learning as a retention tool. They basically did learning budgets so that you feel you had some progression and you wouldn't churn as an employee.

You would not leave the company because you felt the company took care of you in one way or another, but it was optional and it was a thing you could opt into. And too many companies are approaching this in the same way when actually they now have a responsibility. They have a responsibility to the employee. And that is what I can see actually to the person you were talking about. I can see how that can suddenly, when the leader's trying to lead and is doing it maybe slightly too aggressively, can lead to some fear of exactly, it's like, okay, now I'm being asked to like absolutely do this. My job might depend on it. And that isn't— I'm not saying that is a good thing at all. I'm saying that is definitely something that needs to be managed better. But between the two fears, I would want to be in the position where I'm being forced to use it more and being set up to be successful in the future.

Given access to all the tools, having the tools paid for you, all those things, learning on the company's dime, essentially.

100%. And I think that is so important. The real problem is the leaders that are not currently making this clear. They are actually the ones that are setting their team up for failure over the next few years. And it is a responsibility, and I don't think enough leaders are taking that responsibility seriously at the moment.

So let's talk about how to actually implement AI in our organizations. You actually say it's not really up to IT only to implement AI. It's actually an HR problem. Talk to us about that.

Yeah, it's actually not even just only. This comes back to this idea that only 5 or 10% of what we do is help people use the tools. The tools are easy. Deciding to use ChatGPT or Claude or Rebol, like, okay, there is a technology discussion at some point. But what you do afterwards is what really matters. If you just roll out the tools, you can't be surprised that you're getting very little back. The tools don't adopt themselves. There was a friend that told me the other day, it's kind of like giving someone a Boeing 737 and expecting them to be able to pilot the plane without any guidance. It's like, sure, it's like a really advanced machine, you can do a bunch of things. But if you don't tell the person how to use it, they can't do anything. One of the more important signals and important things that we try to help the company with is to help them realize and move the responsibility of AI from the CTO to the CHRO, or at least someone who is looking at the people side of things so that they understand that this is not a technology question.

And that is the number one thing. They shouldn't be looking— this is hard. And I say this, I am a software engineer by background. I was CTO at Super Awesome. Like this is counter to a lot of what I would have said before. But generative AI is a technology unlike anything we have had before. And the reason this is so difficult is that it is technology, which means that all of our pattern matching puts it with the CTO, puts it in the technology bucket. And so by default, it is the CTO's technology, the CTO's responsibility. But the impact is on the people. And so once the technology is rolled out, the real work starts. And that real work is people work, it's process work, it's business work. And it's no longer the responsibility of the CTO. And the biggest signal we see is when we are able to shift that responsibility from the CTO to the CHRO, we know we have at least had a real impact on the business. At that point, we know that the business will, everyone will have different pace of development going through, But that's the number one signal that we would look to.

When AI first, you know, came out more broadly where like companies started to really use it, there was, um, a trend of kind of using AI to actually cut workforce and to lay people off and basically replace workers. And a lot of companies like Klarna, for example, uh, were kind of first to lead that and it actually failed. Uh, whereas other companies like IBM decided to reskill and upskill their employees instead using training similar to what Mindstone does, I imagine, right? So, talk to us about the right way and the wrong way to think about rolling AI out to your organization, and what are some of the key things that we need to, like key challenges or obstacles that we need to look out for?

So, it's a very, very good question. And I think even outside of IBM, IKEA was a really good one. So, the, It is actually comes back to a very similar starting point, which is when it, when AI is the responsibility of the CTO, technology has been for years and years and years about automation. And AI has been about automation for years. When it is the responsibility of the CTO, the focus ends up being on automating work away. And a side effect of automating the work away, if you're not being conscious about what the results are afterwards, is that the work disappears and you haven't figured out what other work needs to happen, which means redundancies. It also happens at board level. And this is actually one of the biggest things we are not doing enough of yet. We get to work with the executive teams a lot. We don't do enough, uh, work with boards yet. And the problem with that is that you still have the same thing, which is that we'll work with the executive team. The executive team understands there is a whole other side to generative AI and for I got to mention this, like the flip side of automation is augmentation, is how do you think about the enhancement, the improvement in quality, the moving the top line, helping the company grow faster, helping it make more revenue, do better work, all of those things.

That is actually where a lot of the value in generative AI lies, but it is not historically where technology has had the biggest lever. So we often work with an executive team, and even when the executive team gets it after a while. And I had one of the more memorable experiences I've had this year is training the executive team of one of the Fortune 50s and really feeling that we pierced through. They really got it. Was like 3 hours into a session, they were really excited, all of the stuff we could do. And then the CEO stands up and is like, great. I've been told to find $1 billion to cut out of our budget. Where do we get it? Which was good. I mean, it's a win for the business, but it was exactly the opposite of where I was trying to put the emphasis. Like, don't just think about where you're going to gain the costs. Think about how this can double your growth rate. How can you find $3 billion in additional revenue in order to maintain where you're at or move people around so you don't end up always focusing on the cost reduction?

This is a real competence problem at the moment. People don't know what they don't know.

Yeah.

So they focus on cost optimization instead of augmentation. This is where this drives to Klarna, that they The bit when, and I don't, and Klarna are doing some awesome stuff, right? So this is, the reality is—

Yeah, no shade on Klarna.

Exactly. People get, first movers get things wrong. We got so many things wrong. Like honestly, I have this rule, like if you look back at, especially in AI, if you look back at how you were operating a year ago and you're not ashamed of—

How you were operating 2 weeks ago.

Exactly. If you're not ashamed of like how you were operating 6 months, a year ago, 2 weeks ago, then you're not moving fast enough. And so they were one of the first ones to move. They moved maybe a little bit too quickly. And I think the thing I do think that was maybe a failure of the way they could have led is that they led with the redundancies. And so they decided to automate customer support. And a whole bunch of people lost their jobs. And what I do think and what I would love people to walk away with is the case that IKEA made, which was completely the opposite. IKEA took their customer support employees and basically decided to put them through training to make them all interior design consultants. So what they did is they looked at what does customer support or customer success look like for IKEA? And they said, we now have this technology. That can make everything better? What does customer success look like if it were twice as good? In the case of IKEA, that meant everyone now has an interior design consultant. And so they thought, we're gonna elevate this experience.

And I think this is the thing that people need to take away, which is you can think about a problem two ways at any point. You can try and figure out how do I get to the same level at half the cost? Or how do I get, keep my cost, but make it twice as good? And I want more people to focus on that second bit for two reasons. One is because it actually leads to people keeping their jobs and everyone getting a better outcome. Two, because it's underinvested in and people would be surprised how much easier it is to get to that point. It's just that we're not used to thinking about it.

Mm-hmm. And the companies that are gonna stay and grow and thrive are gonna be thinking about how to actually make their stuff better.

100%.

Yeah. So what about the fact that AI is not free? It costs money. I, I, uh, me and Kate were kind of doing the math, like, cuz I was, I was, Kate's my business partner because I was working so much on Rebel and I was like, man, this is gonna reduce the cost so much. And she's like, well, according to like how much we spent, It's about $1,000 a week right now. And I was like, well, that's kind of worth it for everything that I'm building, you know, for me as a CEO. But we need to kind of put guardrails on other employees because we can't have everybody spending, you know, let's say $500 a week or whatever it ends up being. So talk to us about how we can navigate kind of how much spend AI costs and like how we should be thinking about that.

Yeah. So this is, A very live topic for us as well. For context, right now on the software engineering side, we spend $5,000 a month per developer. And on the non-software engineering side, we're spending, it depends on the employee. Right now we're spending about $1,000 or $1,500 a month, but we, that is because we are not yet as effective as we want to be. I think it is really important to look at where are we at right now and say over the next 6 to 9 months and what long-term is going to be where we end up. So right now I would look at the API spend as a KPI, weirdly enough, to optimize for. I would say as a business, and we actually tell this to all of the customers we work with, when your API bill goes up, you should celebrate that across the organization. You should have an incentive system where it is the person that racks up $500 in a week should be put on a pedestal because they have been leaning forward, they have been trying things, and the results that you will get as a business from the fact your employees are trying it and really leaning in are going to be so much more valuable than what that costs that right now it's actually a thing you really should be optimizing for.

Internally, we have a target and we kind of think about it the other way around as well. They actually, when you do $500 a week, you're probably, it'll be interesting to see how much you can push that. Most people will naturally, once they've got kind of Rebel in place in different ways, they will naturally top out at about $20 to $30 a day at a maximum. And that's when you start to really use it. Um, so you're using it really well.

Yeah, I, I think I misspoke actually. I think for in a week I, I did like about $300 or something. And then it was like, she was like, it's about $1,000 a month. Yeah. And actually it was like, that's really good. I mean, for how much I did, $1,000. So I misspoke on that. It was about $200 or it was over $200 for the week.

Yeah.

And she was like, it's about a thou— she was saying like, it's probably gonna be like $1,000 a month. Okay.

So that is then you're at that level of doing it really well. And so short term, that is actually what people that get there tend to get a dramatically higher return. Now you do want to, when you start rolling this out for 100,000 people, obviously you want to start looking at, okay, there might be a few people that end up spending much more than what they're making back. But that's the exception rather than the rule. Um, so short term, I would say optimize for it.

And then also, if you know how to use it well, you don't waste, you don't waste the credits, you know what I mean? Uh, and, and some of it can be pretty cost effective. Like I was looking at, cuz I was, I was building all these apps on Ripplet and then I didn't know that it was charging us. And so again, I told Kate, I'm building this app. She's like, well, be careful because it's charging us. And I was like, oh shoot. And I go and look and it was like, I built 2 apps and it was like $10 or something. And I was like, oh, this is fine. I can make as many apps as I want, you know? So it's not that expensive, but to your point, rolling it out to 100,000 people.

Yeah, so to give you a direct example, the other day, so I am having a conversation with a few investors here and there. And so as they were, starting to become more serious, I thought, okay, let's, let's be a little bit more thoughtful about this. Let's figure out who would be the 10 investors in the world that I really need to go and have a chat with rather than just randomly waiting for the few that I happen to just talk to. Normally, previously in Super Awesome, like the process would have been manually researching 200 investors and or actually manually researching 500 of them to then whittle it down to 200 that might be interesting to speak to, and then the 100 that you actually get a path into, and then basically try and have a whole bunch of conversations to go through. That would be 1 to 2 weeks worth of work of just doing the research, making sure you're prepared, going through things and figuring out who goes and introduces you. I did this last week with Rebel. And I think the conversation cost me about something like that. One conversation cost me about $25.

And Rebel actually shows you per conversation what you, what it costs you to go through. But in that time, it had researched 1,000 different investors, put everything into an Excel sheet, rated every single one of the investors based on how Mindstone operates, what their thesis is, crawled through all my own emails, figured out how many of those people have I already talked to? Do I know? Am I connected to them on LinkedIn? Re-ranked them based on thesis overlap and proximity to being able to get either already knowing them or having a direct introduction, and then just gave me the top 10 that I should talk to. Honestly, 2 weeks of my work, weirdly enough, is worth more than $25.

Yeah, that's crazy. Or like, you know, that would have been a $30,000 consultant who wouldn't have had the context or written it in your voice or, you know, so that's incredible.

And so that's kind of short term. So I think like that is, I think, a really important thing for companies to understand, which is short term, If your problem is you're spending too much on the API, actually let me revert that. It is easy to think your problem is going to be that you're going to spend too much money on the API. I have only seen that by some CEO or CTO that knows what they're doing. They let it loose, they're like, oh, I just spent $200 in a week. And then they extrapolate thinking, okay, if I roll this out across my company, every person is going to now do $200 a week. They are not going to be doing $200 a week because they're not you. They have to go through their own AI adoption journey. They will start with spending $1 a day. And actually, your problem is going to be that you're going to have to show them how to make much more, how to take much more advantage of it. That is your real problem. If you ever get to the point where you have that API bill that's so high, it just means you're the most AI-forward organization in the world and you will have doubled, tripled your your numbers.

Now, longer term, I have actually recently changed my mind on how this will play out. So I used to think that for a very long time, it would make sense to just use the frontier models, the best models that exist, the most expensive ones, which are the ones that we'd be using in Rebol as well. And well, that frontier keeps going up, actually. The most expensive models there get better and better, but they're also getting more expensive. Now, when you start looking at the trend at a per-intelligence level, so for the same quality intelligence today, you're paying about a hundredth of the price of what you would've paid last year, like a hundredth. And what is happening is right now, literally the period we're in right now, we are for the first time at a point where Many of the frontier models are at the same, if not higher level than intellectually, than most human beings. So what that means is the reality is that Rebol often knows better than I do. And I have to really double-check myself every time that it comes to a conclusion that's different to my own. I really have to figure out, wait, how did you get there?

Explain it. Let's walk through this step by step. Let's see it. And most of the time it was right and I got something wrong. Now, the reason this is important is that actually we're about to get to a point where, and now it'd be interesting to see just the next generation of models, and we're talking 2 months from now.

Oh gosh.

I know, it's going so fast. The next generation of models, I think for the first time, we'll get to a point where most people won't notice the difference anymore. Because unless you have, for the last 3 years, built up systems that go beyond your own intellectual ability, you won't see the difference between those 2 levels of intelligence because the model is just more intelligent than we are. And so you no longer absorb the difference. You won't be able to take advantage of it. Ultimately, you have a question. You reason it through, it gives you an answer. If you don't have the systems in place to try and figure out, was it the right answer? Was it not? How do I think about using this intellectual capability? You just can't use it. Which then means that for the majority of people, this approximate level of intelligence of where we are now, and it's going to become the harnesses around it, the agents around it are going to get better, but the intelligence level underneath in a year from now is going to be at a hundredth of the price. So right now you might, for this level of intelligence, you might be spending $200 a week.

Great. That means that in a year from now, you'll spend $2 per week for this level of intelligence. Now you might be able, as you now, 'cause you're starting earlier than most, you're definitely on the early adopter curve. In a year from now, you will have built systems that will allow you to take advantage of a higher level model. So you might not be spending $2 a week, You might be spending $20 a week, $50 a week. It's unlikely that you'll still be spending $200 a week. If you do, however, you'll be the most AI-powered, AI-first podcast network in the world, and you'll have taken over everything.

Yeah. Let's talk about right now. I know things are moving really fast, but right now we still need humans in the loop. We can't just let AI run wild. It's the reason why in Rebel, you know, it doesn't automatically send emails, it drafts, and then there's a human in the loop. So what would you say in 2026 are the things that you won't be automating with AI?

It's a good question, and I'd, I'd love to just revert to the human in the loop bit. Yeah. Because actually the framework that, that we prefer is human at the helm. Mm. Which is a different thing. But I think it's a really important nuance difference, which is that human in the loop makes the human the bottleneck because it somehow means that it's kind of like the manager that asks the employees to run everything through them. You're always becoming the bottleneck, which means that the maximum level you can get to is always how much is going through you, right?

Yeah.

It's the same of autonomous cars, actually. The autonomous car analogy here is not a bad one at all, which is that If you still need to be in the driver's seat and you're still responsible for, and you still have to have your hands on the wheel and do everything, there's only so much benefit you're going to get from an autonomous car. Now, the reason human at the helm is different is the idea that you as the human take responsibility and you're accountable for what happens with the AIs that you put in motion. And then you are the one that decides what level of autonomy do you give them. You tell the, you set the context in which your employees understand what they're allowed to do and what they're not allowed to do. In some cases, you might tell them, I'm happy with you sending emails. In some cases, you're not. Email sending is an interesting one because it's a hard one to get past.

Yeah, because it's your, it's you.

Yeah, exactly. I'm starting, like there are now one or two scenarios where I'm starting to get to the point and I'm explicitly telling it, you can send this email.

I do it from my internal Slack messages.

Yes, exactly.

So I left, I'm like, tell Ebi what to do, you know, tell so-and-so what to do.

This is, yeah, it's so interesting.

It's like internal Slack messages. I'm fine, just do it.

Yeah, so I have the same thing. Literally just coming over here, actually, I had a big piece and I I had to come here, but my agent was still working and I kind of had just like, okay, well, when you're done with this work, send the result to Hannah and like do it in whatever way. And then I just walked away from the machine. And so the way that we're trying to instruct this now is that if you're very explicit with Rebol, it will go and do it. If you tell it, send the Slack message, or send the email, then it will go and send the email. If you tell it draft and then somehow the— and this is the thing where it's hard— somehow the AI then decides, oh, well, it asked me to draft, but I'm gonna send it anyways. Which actually, this is what has happened with like OpenClaw and other things. And every time that you hear this horror story, it's like, well, you told me to do this thing and I just decided to wipe your whole hard disk. It's like your computer is gone. It's like, Oh wait, you didn't want everything out?

Oh, that was terrible. I shouldn't have done that. That's the bit where we come in, where, where Rebol as a system should take over. It's like, hey, the user asked this thing, asked to send an email. Wiping your hard disk doesn't seem like an action that should happen when the user is trying to send an email. We might want to ask the user if they're okay with this. The same if you're saying, draft me an email, and the AI tries to send the email, it should ask you, hey, okay, they asked to draft the email, not to send it. So the way we're thinking about it is you need to have multiple systems around. This is where the AI operating system comes in, where Rebol really, I'd say this is probably one of the cores of Rebol is that it doesn't just leave fully autonomously the agent to go and do, which is when it can really mess up. But it has multiple loops around it to verify based on the action you asked it, Is this actually an appropriate action to take? But ultimately, the responsibility is on the human. When you tell it to send a Slack message and that Slack message is off, the thing that I am really keen, and I keep hammering with our own team, the thing that I absolutely do not want to see is people blaming Rebol, blaming the agent.

It's like, well, Rebol wrote this message in a bad way, sorry. It's like, no, you told it to send the message, you didn't check the message, you take responsibility for it.

Human at the helm.

Exactly. The same for an employee. If you have an employee that does something badly, if you're the— anyone's familiar with building companies, the last thing you want is this kind of blame culture where you say, oh, well, sorry, really sorry for what happened there. It's like I had this employee in my team and they just really messed that one up. It's like, no, it's like you as a leader, are responsible for what happens in that team. And if you don't take that responsibility, your team will always be limited to routing everything through you. And that's why I think it's such an important bit, both in terms of how we teach people to interact with AI, it's they are accountable for it, and also to make sure they're not becoming the bottleneck. And this is when we think about setting people up for success in the future, constantly thinking through what is the system I want to build that makes me comfortable giving the AI as much leeway as I am willing to give it. That's probably one of the biggest things that we try and help people understand. And not everyone will draw the limit in the same way, which is fine.

Is there anything in 2026 right now that you absolutely will not delegate to AI?

The human connection is a really, really big piece. Here's a reason why we're doing this in person. There's a reason why almost every single executive briefing session that we do is in person. It's because a lot of this AI transformation is actually emotional. And emotional is not something that AI is, I mean, it can put the words in place, but it can't really make someone feel better about it, or at least it's harder. It's a fundamentally human experience that we're going through. And then the other bit is agency. I won't let AI just self-execute on plans within the business without me. So there are ways technically that I could now go and set it up to autonomously start executing on basically whatever it thinks needs to get done. I am spending my time figuring out where do I apply my own agency and how do I build the system in a way that I maintain agency over what happens, and then take the most advantage of the AI from within that. But so, not letting it act without me having asked. And then second, having that human connection and that human experience is where I would put all of my time.

And hopefully, I mean, this is an interesting one because it's probably counter to, definitely counter to how people would perceive me externally. It's so interesting because I'm so absorbed in some of the technology.

Yeah.

The number, one of the things that shocked me at the time when we were building Super Awesome was interesting. People sometimes think of me as somewhat robotic because I spend a lot of time with the technology and I'm very logical, but I care a lot about the people that I work with. And so one of the things that using AI allows me to do is try and spend more time actively with people and making sure that I make more time for exactly that piece. Yeah. So human connection.

Okay. So as we close out this interview, 'cause we're, we've been talking for a long time, honestly, I feel like we need a round 2 because as you, you saw the prep brief, I have 50 more questions to ask you, more detailed. I think that belongs more in like a part 2. So we'll definitely have to set that up. But, um, I wanna talk about tool tourism, right? So, so this is something I think you might have coined where people are just trying every tool. Uh, what's the problem with tool tourism and how should we think about picking tools that we work with?

I wouldn't quite frame it as a problem, but there's def— there's definitely—

it's a situation.

Exactly. It's definitely things that you, you need to watch out for. So I'd say on balance, Tool tourism, meaning that you're actually interacting with the tools, already puts you ahead of most people who are not even interacting with them, right? So that's what I mean with it's not necessarily a problem. It is definitely, it would prevent you from getting to the real return on the investment of the time that you're putting in. And I think this has also changed in the last year. Which is that the technology in the last year has really turned a corner where it is now able to produce real work. If you just sit at the level of tourism, you're just superficially trying a tool, you don't really try and go deeper, you run the risk of never going deep with any of the tools and then never actually getting to the behavior change. And if you're superficially looking at ChatGPT and then Claude and then Copilot and Gemini and you're kind of thinking, oh yeah, I try all of them out and I kind of do I do these things and I compared them between each other. Great. But if you didn't go deep with any of them, you actually never got to the result or the return with any of them.

Yeah.

So what is important is that you go deep with a few. Now that doesn't mean it's bad to try the others. It's always good sometimes to compare and contrast, but don't let it take away from you being able to go deep with one because otherwise you just won't understand what's possible. It's kind of like, If you had 10 different operating systems and you tried all of them, like, I mean, right now we really only have Mac, Windows, and Linux, but imagine you had 10 of them. If you change that operating system every week, you would never feel comfortable with your machine.

Yeah, that's true.

And at some point, you just need the machine to sit in the background. It's there to help you do work. So whether it's ChatGPT or Claude or something else, or Rebol becomes the operating system that you live in, It is important that you give it enough time to really go deep.

Well, you've toured a lot of tools, so I wanna end with a quickfire segment on the best AI tools. Okay, so best AI tool for building presentations quickly?

Canva.

Best AI tool for meeting notes?

Fireflies at the moment.

Best AI tool for research?

Gemini's Deep Research.

Best AI tool for brainstorming new ideas?

ChatGPT.

Best AI tool for thinking and communicating clearly?

Claude.

Best AI tool for graphics or video clips?

Tie between Klinger and v0.3 or 3.1.

Best AI tool for building apps without coding skills?

Replit.

I agree. Best AI tool for coding with coding skills?

Droid.

Best AI tool for learning faster?

ChatGPT.

The AI tool that more people should know about?

Rebol.

I agree. The AI tool most people are currently using wrong?

All AI assistants.

All AI? I was gonna say—

Well, all AI assistants from Copilot to ChatGPT to Claude to— yeah. Number one problem is people use them as search engine replacements and they just don't work that way.

Yeah. Okay. So last bit is about Claude versus ChatGPT. So what does Claude do better and what does ChatGPT do better?

So Claude makes it easier to talk to it like a human in a way. It actually talks to you in a more humane way. Its writing is much less mechanical. It's more elegant in a way. It is more human. Yeah, I agree. ChatGPT is really strong at reasoning though. And so when you need to crunch through a logical set of steps, ChatGPT can be better at it. Also, ChatGPT was first and their ecosystem is more evolved, although Anthropic with Claude is starting to make things—

Yeah, it's kind of really, yeah.

It's really getting there, but for a very long time and still, One of the features that ChatGPT has that Claude does not, which is fundamental, which we didn't even get to on the weekend, and I just realized we should have done that, is the ability— so ChatGPT has a canvas which allows you, when you're brainstorming, you can put it in a document and then you can annotate the document, but you can basically go back and forth with the AI in the document. It's almost like having a Google Doc, but with AI. Using change suggestions and comments as it goes along. And when you're thinking about brainstorming something, often my flow would be actually start some of it in ChatGPT, go through, prepare the document, put it back into Claude when I want it to sound more human. But so that would be, yeah, another way that they compare and contrast.

If somebody is listening today, they haven't really started building, maybe they've used ChatGPT loosely 'cause everybody has at this point, but maybe they haven't really like built any systems yet. Would you say ChatGPT or Claude?

Claude today.

Yeah. Yeah. Because Rebel uses Claude as one of the main, and I, I feel the difference in quality is like so much better that I have stopped using ChatGPT nearly as much. Okay. If you had to delete one forever, ChatGPT or Claude?

Definitely ChatGPT.

There we go. Claude is the winner. Joshua, this has been such an incredible interview. I end my show with two questions I ask all of my guests. Uh, the first one is, what is one actionable thing our young improfiters can do today to become more profitable tomorrow?

That, I mean, that is the perfect question right now for, based on this conversation, is just whatever you do, lean in on giving yourself the tools to understand how to use AI, uh, whether you do that with us at Mindstone taking the competency program, or do it yourself, or talk to your company about what they can make available to give you the space. Whatever you do, make sure that every week you spend some time figuring out how deep this rabbit hole really goes. And what would you say your secret is to profiting in life, especially with all these changes going on and right now, it's genuinely that I feel for the first time in my life that I'm living the future, which is a very weird position to be in. But that's why we're having this conversation today, is that I feel that what I'm able to do now, most people in the world will be able to do in the next 2 or 3 years. And everyone, interesting enough, everyone has a choice and they can do the same. It's not even that I'm special in one way or another. It's just that I leaned in before anybody else did.

And so that is the secret sauce today.

Yeah. I feel like just so lucky that I was able to go on the AI retreat. I told Kate, like, I feel like it was fate that we were there and now we're going to take 60 employees on this journey with Mindstone to teach them how to use Rebble and to, you know, understand AI. And I just feel like it's really gonna help my business stay competitive, thrive. It made me realize how at risk we are if we don't do this. And so I'm just so thankful that we've met and I'm grateful that now my audience gets to hear everything about Mindstone. Where can everybody learn more about you and Mindstone and everything that you do?

So very simple, go to mindstone.com. Easy. And so we have the AI competency program. Everyone can enroll, whether it's individuals, small teams, and if you're a company that is interested in kind of taking your whole company through it, there's a contact form. So we basically have it for individuals and for companies, and then we can have a conversation, figure out how far we can go.

Amazing. I'll put all those links in the show notes. Joshua, again, it was such a pleasure.

Thank you very much for having me.

A huge thank you to Joshua for pulling back the curtain on what's truly possible with AI right now. That was definitely the most practical and executable episode that I've ever had on AI. This conversation genuinely shifted something for me, and I hope it did for you too. One of my favorite insights is that most people are using AI completely wrong. They're using tools like ChatGPT and Claude like search engines. Joshua challenged that mindset. Instead of asking AI for quick answers, treat it like a smart teammate. Give it context. Ask it to ask you questions first before giving you an answer. When AI understands the full situation, the quality of your ideas, strategies, and decisions improve dramatically. Another big takeaway is finding the coding equivalent in your job. Every profession has a repetitive task that eats up time. It could be updating spreadsheets, doing research, drafting emails, or organizing information. Joshua's advice is simple. Identify that task and start experimenting with AI to handle it. When you automate the repetitive work, you free up time to think strategically and operate at a much higher level. And finally, start building systems that work for you. Joshua talked about creating skills, which are repeatable workflows that AI can execute.

Instead of doing the same task over and over, teach AI the process once, whether it's research, follow-ups, or content prep, hours of work can suddenly just happen in minutes. That's when AI stops being a novelty and becomes a real competitive advantage. The opportunity with AI right now is massive, but it's not gonna stay open forever. The entrepreneurs who embrace this fully are gonna get a moat around them. They're gonna be untouchable. I want that to be you guys. You're already ahead of the curve by just listening to this episode. Now go take action. If this episode planted a seed in you and you feel more motivated to use AI than ever before, then water that seed by sharing it with somebody in your life who also should be interested in AI. Your only homework today is to pick one thing from this episode and actually try it this week. That can be talking to AI instead of chatting with AI. That can be trying to use a new tool like Claude or any of the other tools that we mentioned in the episode. And then come tell us how it went on Instagram @YAPwithHala or LinkedIn by searching for my name.

It's Hala Taha. And if this episode was worth your time, a 5-star review on Apple Podcast goes a long way. Those reviews help us reach more ambitious people just like you, and they mean the world to me. I personally go check to see my new reviews every single day. I'm obsessed with checking my new reviews. So please drop us a 5-star review. And if you guys wanna watch the full conversation, We did record this episode in person, so you can find it on YouTube or Spotify Video. Just search up Young and Profiting Podcast on YouTube and on Spotify. You can just watch it as a video. This is your host, Hala Taha, AKA the Podcast Princess, signing off.