People that are winning with AI are the ones that are not watching videos, but actually they're building. It's no longer time for learning, it's time for building. More and more people are embracing vibe coding. Sometimes you need to forget about everything, how you've been doing things and literally just like start from scratch with these magical hammers that we all should be using every day. Today we're joined by Iwo Schappar, a serial entrepreneur who's been building infrastructure for the future of work. From co-founding remotely how, years before the pandemic, and now creating the AI maturity index to help companies measure their readiness for tomorrow's workplace. Iwo is one of those rare founders who sees around corners for anyone else does. Welcome to the podcast. It's great to be here. Thank you for having me. Well, I follow all of your travels and the work that you're doing, whether you're in Dubai or Warsaw or somewhere around the world, you're really helping people navigate change, whether it was a change to remote work or now this new technology. I want to start with the AI maturity index. I've got over 250,000 data points across 75 countries and you're working with Harvard and some of the brightest minds. What are you seeing like the reality versus the hype? You and I and many others who follow this space constantly see algorithms telling us that you're all way behind and what are you seeing? I think the good way to start is that this is the end of the beginning. So although it's been almost three or two and a half years since the Chudge GPT moment, most of the market is still in the very early days. We have the AI bubble, people that are using AI every day, experimenting, building solutions. And this is this tiny, tiny, tiny niche. Then we have a broader bubble that sometimes can be mainstream people using Chudge GPT. But first, I would say simple task like write me this email, review this document kind of thing. And then most people still didn't build up the habits to experiment AI with more and more use cases that are truly making an impact on their day-to-day productivity. So with the A maturity index, we are uncovering some really fascinating insights. And we've been digging into the data on how the top 1% of the AI users compared to the bottom 50%. But I'm more than happy to share some stats. So first of all, the top 1% saves 31 hours per week thanks to AI, while the bottom 50% saves only six hours. So it's like over five times performance gap, which is really mind blowing. If you look at all the tools that are available there, and actually the fact that it will save you time, like of course, like in the beginning, you need to learn a bit, you need to build up these habits. But the payoff is really coming very fast, right? The second thing that we see is that the top 1%, 94% of these people use AI multiple times a day. While the bottom 50%, only 12% of them uses AI multiple times per day. So here you have eight times the frequency gap, right? And I can go on with some data, but I think to me, the biggest mind blowing stat was when we looked at the different dimensions. So we're looking at productivity, we're looking at collaboration, creativity and decision making. And the largest gap on how people are using AI, the top 1% versus the bottom 50%, is on decision making. And the gap is almost 13 times, right? It's mind blowing, looking at the fact that AI is such a great advisor, such a great advisor to help you analyze the data, to help you also just like your sparring partner, your mentor, unless you give it a contact. Like you need to give it a contact, right? And I think this is where still there's like a huge learning curve on providing this context, both on the individual but also company basis. So TLDR is we're just starting, things are moving very fast and we really encourage everyone to experiment, try because unfortunately, otherwise you will be left behind by the ones that are doing this every day. It's interesting you say that I see a lot of videos. Traditionally, corporations, when I used to work in big tech, that you were going to learn something, they'd roll out a LinkedIn learning video or something. And what I've come to believe to be true is the only way to really learn AI is to kind of dig in and experiment. Like it's messy, you have to prompt and learn and play with a lot of the different tools and you need the time to do that. When you talk to people, let's say they're in that bottom 50% cohort that you mentioned and they feel left behind. Like, hey, I know that I need to be learning this technology. I know it can be helpful for all the reasons that you mentioned. What are a couple of things, like really tactical things that are important for them to think about to start doing today? Yeah, so I really like this role that any task that you do twice should get automated. Just take a pen and paper and start writing down the things that you're doing. And if you notice that something is like, oh, I will be doing it again in a week or even tomorrow, figure out if there is a way to either try to automate the whole process. Tough, right? But that's kind of like a North Star. But maybe see if certain aspects of it can be automated so then you can start digging into this more and more. And it's like a very simple process. So first, you should capture the process. So you can screen record it with LUM or, I don't know, Slack video, whatever. And then talk about this process. Talk what you're doing, describe the actions, et cetera. Then dump this transcript into any Lamp, OpenAI, Gemini, Cloud, whatever. So that's kind of like a step two. Then step three is that you're codifying this. So ask AI to generate an SOP, so a standard operating procedure for this, and put it somewhere in Notion or any of the knowledge database that you have. And then third, if there are certain aspects that can be purely automated, there are tools like Zepure, Maker, dot com, et cetera, where you can connect the dots and start really not just talking about automation with AI, but actually start doing this. So really, if something is happening more than once, think about how you can automate it or how you can augment it. And the second very tactical advice is that, same start, like start writing down the things that you're doing. And then what you can do is you can group them into three buckets. The first bucket is automation. And then, of course, you can follow the process I just mentioned. Then the second one is augmentation. So those are the tasks where you actually need AI as you're like a co-pilot assistant kind of a thing, right? But you're guiding the conversation. And then the third group of tasks, you're like, oh, no, this is just human only. This is where I'm working by myself or with other humans, right? And this will give you a first overview of where really AI can help and what needs to be done in order to reach this state when either it's on fully autopilot, automation, or you're working hand-by-hand with AI. But you already understand these different areas and different opportunities, right? Because there's a lot, like you said, like we see a lot in social media. This is possible. That is possible. New things all the time. And we're getting overwhelmed. So going into basics and really just like nailing down what you're doing and then figuring out how this process can become AI first is an absolute must. And maybe last but not least, people that are winning with AI are the ones that are not watching videos on how to do stuff, but actually they're building. So I really like this phrase that it's no longer time for learning, it's time for building. More and more people are embracing vibe coding, meaning like you don't need to be a software engineer, you can start building software by just typing natural language. So really in 2025, the limitations is just like our imagination. Of course, there are still certain technical limitations, but for a lot of use cases, it's already possible. It's just a matter of building it one and then testing and improving, iterating, iterating, iterating. Because with AI, especially if we're talking about the multi-agent systems, it's a living system. Like you need to keep iterating, you need to keep adjusting. But it's fun. It's a totally different experience of working. And I strongly encourage everyone to really, sometimes you need to forget about everything, how you've been doing things and literally just start from scratch with these magical hammers that we all should be using every day. It's interesting, one of the things we're all bad at is writing stuff down. And you said something that the ability to get a transcript, the ability to take audit AI or any recording like we're doing right now, and just talk to it. I go for walks and sometimes record notes and stuff and then come back and take those transcripts. And so finding areas in your workflow to get stuff out of your head, to talk about tasks, because it doesn't cost anything to record an hour and have it transcribed like it used to be, you had to take a recording and maybe get a freelancer to do the transcription. Now the technology, the AI technology can transcribe everything that you say in real time and make it helpful and actionable. So if you just sat down and like you said, screen recorded or even used a transcription like Otter or whatever, you can get that information out of your head and now start working with it in real time. If you think about not just even getting your thoughts out there transcribing, but actually getting instant feedback or advice or a brainstorming partner, I think like anyone listening to us that is like, ah, I don't know how to start, like how to get to next level, just ask AI. Like, hey, this is who I am. This is what is my job role, what is the company doing, like how can I benefit from AI? And also one important thing, if you don't have a paid subscription to any of the LLMs, this is probably the best investment in your professional growth that you can ever do. Like I'm not selling anything, I'm not getting paid by any of them, but I truly encourage everyone just to even try for a month, because the models that you're getting access to are like way better and a lot of different functionalities, so it's just a foot for thought if you're hesitant. It's interesting, we talk about work use cases and productivity, and I was talking to a person at one of my hackathon events, and she was telling me how her daughter had gone through a breakup and she wanted to give advice to her daughter. She literally went to the LLM, talked to it, and it gave her some insights and advice on how best to work with her daughter during this challenging time. So I always encourage people when you think of AI as a thought partner or just anywhere that you might need help, try it out, because nine times out of 10 or 10 times out of 10, you're going to be surprised at the insights that it helps you think through a problem or think through a recommendation on how to approach something, whether it's at work or at home. You mentioned vibe coding. I'm not a coder, I've never hit compile. I've worked in technology for all of my career. And I used about three or four months ago, I use lovable. It's one of these no code platforms. And I built three apps. Now look, these apps are not production ready. I'm not going to go be a multi-billionaire because I created the next Facebook ad bit. But I was able to create an application by typing a couple of prompts, a couple of what I wanted into the SAS tool, and it created things and it took care of the web hosting. So I didn't have to set up or configure or go to go daddy and do any of that. It did the front end. It created the back end, which is all the database and stuff like that. And it even did email authenticate like these very these things that used to be pretty challenging to go and do, even if you weren't technical. And it created a sign up app for an event that I did. And it only took five prompts to do. And when I think about vibe coding, the word coding, it just seems offsetting to people that I don't know how to code. So that's not for me. What are you seeing as non technical business people or non technical people start adopting things like that? So I would start with with a pain point or a use case where in the ideal world, you would have a hundred thousand dollars to pay an agency to build you something internally, right? That you would make your life easier, better, right? And really dream big, dream big on automating something or getting rid of your subscription for your CRL like anything that is really like to change. And this is a starting point. We've already seen a lot of people building real businesses by vibe coding. And we are also seeing a lot of interesting use cases, how businesses are building mini internal tools, smaller ones, bigger ones that are either building end to end workflows between different software that the company is already using, or they're actually trying or already replacing the software that you're buying. Right. So the sea of Microsoft said quite some time ago that SAS is dead. And of course, yes, it's a marketing thing. But if you really think about the ability that the individual, let's say that you're a marketer, right? There's another term right now going popular like vibe marketing, right? And you would like to build this tiny tool for lead generation and analytics, whatever, like you can do it yourself today. And you might save, I don't know, hundred, two hundred dollars per month by just like paying for a subscription, right? Use cases are left and right. My advice is just to start very small with a very small scope with something that is a true pain point. It's something that would really make your life easier. And you would be truly surprised like what you mentioned that people that are attending your hackathons because I've also never been a software engineer before. And I'm vibe coding for like almost a year now. And it's mind blowing. Even some of my code is going to production of a maturity index, which I would never, ever, ever thought that it's going to happen. And maybe last but not least for people that are working somehow with with tech teams are giving them specifications, etc. With vibe coding, you can prototype your idea, not just like doing another document that is describing what you want, but you can actually build a clickable prototype front end, back end and then go to your team and then discuss the idea and discuss how you can scale it. So possibilities are out there. You just need to do it. Everybody keeps going back to AI as a coach or AI as a thought partner. And that's just another example. Hey, I've got this idea. If you go and you use one of, you know, bolt or lovable or there's lots of them out there, just to think through your idea and kind of experiment with it. That's a valuable use of those tools. And I think a lot of times they're sold, they're over-hyped. Hey, you do this and you could have a production app that scales to a million users. And that's just not true. No, no. That's really like once in like million or someone is in software engineer and then vibe coding is just like an hundred X of what they could do. But also just to keep the reality check. It's not that you just do five prompts and then you have it ready. Sometimes I'm getting stuck for like hours because there is some loop that I cannot escape. And maybe small tactical advice, not just for vibe coding. My biggest hack of this year is to tell AI to meditate. It works like a charm because then it clears its mind literally and then restarts. I've had multiple cases when I was like, oh, my God, like, I don't know what to do next. Then may I meditate, breathe, start again with a fresh mind. Game changer. I think we can all take that advice in life in general, whether you're talking to an LLM or not. I want to change topics. You're working to build an AI community of consultants. And you wrote an article for us that we'll put down in the show notes. And there's been a lot of talk about the pressure on companies like McKinsey and Deloitte and others. I've paid millions of dollars when I was in big tech for companies to produce research and decks and stuff like that. And so there'd be some junior consultants, there'd be a partner and they would do a bunch of work to provide research. And a lot of that's starting to collapse a little bit because AI deep research can do a lot of the heavy lifting when it comes to synthesizing that data and the market research and the recommendations. I want to ask, what are you seeing with AI consultants or smaller firms? And where are they making progress? Like, where are those connections happening between people who are experts in their field using this technology called AI and getting clients? I think you've worked with Microsoft and Walmart and even smaller companies. Where are you seeing the magic happen where the gap is between sort of smaller, more nimble consultants and the bigs, like Deloitte and them? So I think like big picture is that AI enables huge SMB market that was never, most cases, open to such consulting on like digital transformation now AI. So there's like a tremendous market of companies that right now must do something and right now it's called AI. So perfect time to offer them your services, ideally not even just services, but productized services. So this number one, tell me more about productized service. I mean, for people that might not be familiar with, give me an example of, hey, I'm a consultant that offers a service and I'm a consultant that offers a productized service. Yes. Let's take the example of doing a discovery with the client, like almost all consulting AI or not AI is to understand where the client is currently, whether they're paying points, gather the data and then give some recommendation. Right. So just pure discovery. If it's just a professional service, then you're doing interviews yourself. You're sending out the survey, you're analyzing the data, et cetera. And right now think about how you can productize it. So then this professional service, you're packaging this as a product. It has a fixed price and then it's a technology. So for instance, you have an agent that is running these interviews, that is analyzing the data. And of course, there is a human touch to kind of like put this all together, but you're leveraging technology in this delivery. And as I said, it's like a fixed cost. So it's also easier to sell it. Right. And I think this is another aspect that has a huge potential because of AI. In the AI world, you can see more and more software companies are shifting from subscription into outcome based pricing. Right. So one of the examples would be Intercom, a company offering chatbot for customer support, where normally you're paying per month for access to their software. What they're experimenting with is your paying per resolved customer ticket, like 99 cents or something. Right. And this shows you that there is a disruption happening on the business model with software that is also going to happen and is happening with professional services. So it will be no longer about the hours. It will be about the outcome. And if you merge it with the product service, it's like, okay, you're paying 5000 for this report and we're doing this in five days. Right. And it's also easier for you to estimate the cost, the time, etc. And it's also easier for you to sell. So those are the two things that are really impacting not just the consulting, but consulting in general, but also coming back to your point, like on the differences and how people with the expertise can bridge this with AI. Like for sure people, they're expert in their areas. They need to become more technical because the real, real challenge is on the actual implementation. So it's no longer that you will provide some nice slides and the strategy and they will be like, oh, nice. Thank you. And then they put it and forget about these slides because here it will be actually okay. So we have the strategy, but now we need to implement it. We need to put this into production. Right. And normally this kind of stuff was handled either by like software houses, technology partners, but right now, because a it's no co-local, vibe coding, all of the terms that are enabling you as an individual to actually go in. You don't need to hire 10 software engineers to join your consulting agency. Right. You might have an advisor or something or just like help build clients yourself. So I think bridging this gap to like be more technical is extremely important. But right now we're talking mid 2025. The biggest challenge for a consultants is to show what are the use cases that the companies can implement, create urgency and start with something small to get the ball running. Right. There's a lot of talk. So how you can convince the client to do something, show certain benchmarks, create urgency, pick and choose one specific use case. And then from this use case, then you can build up to more. You know, it's interesting. I just got finished doing a hackathon for a FinOps company and we had some fun. We created our first AI songs. Everybody created profile pictures just to kind of get warmed up with AI. And then we spent some time using AI to understand pain points, right? And to your point. And then we prioritize them and we walked out of that session having built a prototype. Now it was a, you know, hey, we have a bunch of leads and we want to score those leads and we want to have personalized research as we reach out to those leads. And it was using Google sheets and Gemini and some other things. And it wasn't, I think when, when people think about being technical, the ability to take a problem and solve it with Google sheets and an LLM is possible. We're not building the next sales force, but the team walked out understanding AI better, understanding their use cases, which is extremely valuable to your point. And we actually built a prototype all in a few hours of an event. And I think one of my key learnings to add to your point is that taking the time to go through that process of getting, increasing your AI literacy a little bit, understanding use cases and stuff, and then taking the time to experiment is business critical. It's kind of like when, and I'm sure you've had this, if you work for a company, the mandatory training, oh yeah, yeah, yeah. It makes you sit through hours of this, like the lawyers in HR, like the amount of hours of mandatory training. And I have friends right now who are like, literally don't have time to experiment with the technology because they're in meetings and stuff. Oh, yes. Classic, classic. Like I really liked back in the days when it was all about remote work. A lot of these companies had like a mandate on a weekly basis to have an hour, a few hours every week where you're just learning, exploring new topics, maybe even building. It was like mandatory, like your manager was like, hey, last Friday, have you learned something, et cetera. And I think this is this mindset that will help both companies and individuals really experiment and get the most out of AI. Step one in increasing your AI literacy is to delete seven meetings from your calendar this week. Please do. Yeah, I look at friends and even my wife's calendar and I'm like, I can't even imagine how hard that must be. Even to get your job done, much less start to think about experimenting with this new technology. If you were going to advise somebody, let's just say there's a person out there who may have lost their job in technology or they're saying, hey, I know I need to pivot. It's time. What is something they could do today to start that journey? I think I would do an experiment and be honest with yourself. Where will be your expertise? How your expertise will be needed in five to 10 years? What you currently have? Assume that LLMs will only get better exponential curve. And let's assume that you're coming to a conclusion that, ah, you know what? Like maybe most of the knowledge that I have will be already accessible through an LLM. Then I think that might be the moment when re-skilling a journey would need to kick in. And then the other option, if you're like, no, it's fine. Like actually AI is augmenting what I'm doing right now. I will be able to do more, et cetera, et cetera. Then I would really go back to simple things. I really like the framework called STEP, which stands for segmentation, transition, education and performance. So you're starting with segmenting your tasks into automation, augmentation and human only bucket. Then you're planning that transition. So you're looking at what kind of things you want to start with, what kind of software you already have, what is needed. And then you're starting the transition process. Then you're of course educating yourself all the time. And then you're also measuring this. So I think there's been a lot of talk for many, many years on the measurement of productivity, of output, et cetera. But right now, because you're working with AI and it's tracking a lot of the things, I would strongly encourage everyone to pick and choose certain outputs that are measurable. So then you can actually be like, okay, I've been doing this for a month, three months. This is what has, what has really changed. And either you can showcase this internally or externally. But if we're looking at someone that's starting a new career, it's looking for something completely fresh, I would just say to assume that everything that we know of how the world of work look like, let's completely start from scratch. We have no idea what's coming, but for sure it will be a big disruption, both opportunities as well as challenges. So if you had the dream to open up a coffee shop with flowers or to take pictures of surfers in Hawaii, I think this is the perfect time to go after your dreams, to be honest. I want to go back to having AI be your thought partner. If I was going through that process, I might go and say, Hey, here's a framework called step. You're an expert interviewer in this framework. Ask me questions. I think the biggest unlock I've seen in just my experimentation is ask me any questions needed to complete this task. Because the questions it asks you are actually more powerful to me in thinking through things than actually the output that it gives me. So if you're out there and you're, Hey, I kind of want to think about starting something new, go to an LLM. I think to your point, pay the $20 a month for, I think two Starbucks. I think coffee costs 10 bucks a cup now or something. You know, for two Starbucks a month, get deep research from Open AI or pick your favorite LLM and talk to it. You haven't helped you walk through some of those thought processes to understand your skills and think through what a, And to your point, another important aspect here is to give as much context as possible. It doesn't need to be formatted, just like copy, paste stuff from your LinkedIn, from your portfolio of like anything, your previous work. Because this is the thing that I see almost every day that people are expecting a perfect output if they just input it one sentence. LLM is good, but cannot like guess stuff that you have in your head, right? Or you have it on your, on your like private folders, right? So like dump as much stuff as possible and be very explicit in what you expect to get or to your point to ask you questions based on this context and then to arrive at a certain outcome. Yeah, it's funny when I wrote gig mindset and started working with freelancers. I had to train myself to communicate what was in my brain to another human being because you want to get something done by somebody who's an expert in it. You have to be able to communicate that this technology is not any different. And unfortunately we, as humans, are not good at communicating. That's why I always advocate for record yourself. Use the transcriptions. Just go and talk and babble to a transcriber to get stuff out of your head and then have use the LLM summary technology to kind of organize it into a way that can be communicated to someone. Thank you for your time. I again, anybody who's on social media, especially LinkedIn, I encourage you to follow Iowa and his work, whether it's remote work or the AI work that he's doing. He reminds us that it's fundamentally about reimagining work and where our value is and what it means to be an expert. And that profound change is coming. And so if you're willing to go out there and live, build and think differently, a lot of things are possible. So thanks for joining us. And to everyone out there, keep building the future and stay curious. Thank you so much.
