Who here vibe -coded today? I'm pretty disappointed with that. My hope is within the next time I'm in front of all of you, every single hand goes up.
So, I'm excited to share with you my latest vibe -coded product, which is not working completely this morning. morning. Today will be a talk around where does vibe coding go and what does it look like from today to the moment when every single hand in this room goes up, when every single person is building with AI.
Marcus asked in the last talk who here is a developer. Not that many hands went up.
I studied software engineering. I built a software business for 15 years. I don't actually identify as a developer, but I do identify as a builder.
And in the
last three years, there's now a billion people that are chatting with AI. From my mom to my nephew to my cousins, everybody's chatting with AI.
I think in the next three years, everybody will be building with AI. And not just those of us who come to an AI meetup at night instead of going being at home watching Netflix or doing something else
or enjoying the beautiful weather but I think our family our friends our co -workers if we still have them are all going to be building with AI.
Who here works for themselves as an entrepreneur or solopreneur? Who here works for somebody else?
Who here works for an AI? That was a joke. Marcus is looking at me all serious maybe you're already starting to work for an AI it feels like it yeah yeah it does feel like it I spent most of my day today working for the AI and
taking instructions from the AI so in the past there were the people that could build and then there's the rest of us and for the last 30 years the people that could build are the people that had the power. Engineers got paid more than pretty much every other profession.
Investors who invested in people that could build built a lot of capital and wealth over the last 30 years. The 30 million or so developers in the world basically decided what our experience was going to be of technology over the last couple of decades.
So there was this wall between the people that could build and and the rest of us and that wall is very clearly coming down so
I ran a software company for 15 years and I still couldn't build the technology so I ran everything I raised tens of millions of dollars I hired hundreds of people I sold our product to hundreds of clients and dozens of countries but I couldn't do the one thing that mattered which was build our
product. So over 15 years of building a company successfully, I had a lot of frustration. I had a lot of frustration almost on a daily basis because I didn't feel in control. I didn't
feel empowered. But now I do. And this moment that we're in changes that for all of us. All right.
So 2022, I called ChatGPT a party trick because you'd ask it a question and you You didn't know what response you were going to get. And it was a trick, you used it once or twice and that's it.
And then in 2023, developers started to use it and it was, you know, have it write some code, copy it into a different IDE, a programming environment, fix the code, and it kind of made your job faster, kind of made your job slower.
And then in 2024, the concept of vibe coding started to be introduced and non -developers started to build things as in really ideas and prototypes that's supposed to go all the way across and then so full applications being built by those who have never seen a line of code and then you know applications being being shipped so like we're literally in this moment now where
the most valuable thing is not based on how technical you are it's going to be based on how well you understand the problem. Because in a world where everybody can build, how do we differentiate?
So what changes? So there was this concept of it's not my job.
So for any of us who worked in a company, even me I ran a company, the number of times I said that's not my job, I think we're gonna look back and laugh at that. We would say it's that team's job, the developer's job, but now
Now everything changes, like the idea of a dev backlog, the idea that we're waiting for it to be developed, that will change. Anybody who's worked in a software tech or product company knows kind of the tension the organization feels when we talk about building something new, when a customer asks
for a new feature. So I think that changes now. The power shifts. It's to the person who understands the problem versus the person who can fix the problem.
In the past, and I say the past, the power within a team was the person who could fix the problem. That was the person who got paid the most, that was the person who set the pace, that was the person who set the roadmap. I think that's changed.
The org chart obviously flattens and the way we manage changes.
So what do I tell my nephew as in what do you say to the young person in your life as to what they should be doing? In the past, it was you should learn to code.
You should study science, technology, engineering management, become technical. If you were an entrepreneur, you were told, go get a technical co -founder. I was very involved with the entrepreneur scene here in Portugal and in Canada and New York. And the number one question I would get asked is, where do I find a technical co -founder?
It's been like 20 years. That was the message to anybody who is entrepreneurial but not writing code. I don't hear that question anymore and so in the in the future what becomes
important so taste is important empathy is important the humanities become more important it's a little funny because a lot of us in technology would have laughed privately at these attributes and skills but now they're going to be
the differentiator so I compare vibe coding to playing a video game who was a a gamer here growing up or as a kid okay a lot of you so as as a kid for me i actually never played any video games and i had a lot of judgment towards those that did play video games
and when i look now at what's happening right so there's 3 .4 billion people in the world that play video games and it took 40 years to get there and i believe that within 10 years we're We're going to have a billion people playing with AI to build things. And the game dynamics are very similar between vibe coding and video game.
Are you guys familiar with this simple chart? Anybody guess what word I'm about to say to describe this? Flow state. I'm sure you've heard of flow state.
Flow state is this orange line in the middle. it's when there's a perfect balance and harmony between challenge and skill so when I came to Portugal five years ago trying to speak Portuguese was like right here for me challenge versus my skill actually trying to drive in Portugal was over here but over time now it's over here is it doesn't feel a challenge for me because I've had a lot of experience with it.
So vibe coding is an addictive video game because at least my experience of it is along this line here. And I can keep challenging myself with new projects, new tools, new techniques that map to my skill. It adapts to me in terms of where I'm at.
I spent last year actually living in N8N. N8N was the number one tool I used throughout the year. and now it's cloud code and both create value and everybody here can use both of them but they represent different challenge levels based on your skill.
This is the difference between video games though and vibe coding and every addiction and I'm definitely addicted to AI right now is the addiction to AI leaves me with more, it leaves me feeling more, it leaves me producing more, it leaves me learning more and it's very different than a lot of other addictions in our life.
I actually had a therapy session earlier today, and I used the word addiction, and I was saying to my therapist, like, hey, like, I'm waking up at 3 a .m. I'm not sleeping too well. I'm
socializing a little less. I'm starting to show symptoms of addiction. I'm aware of it, and I brought it up.
And then this was inspired by her, actually, because she said, well, no, addiction has a negative connotation to it, but this is feeding you, and this is leaving you feeling feeling more versus feeling less. So go ahead and be addicted to AI.
So the number of builders in the world I believe is about to go from about 30 million people to a billion people. That's that's what I believe we're about
to witness and experience and I'm very curious about what the impacts of that. I'm using a billion tokens a week right now. It's fortunate that I can can afford to do that.
I can afford to be inefficient and have 10, 12 agents and terminals and my 50 -inch monitor just going and running. Do you guys like this animation? It came up
with it this morning. I'm learning a lot. That's the case for why we're all going to
be building, but what are the implications of that? A lot of disruption also gets married to that right so development especially junior development and a lot has been written about this changes quite a bit a lot of economies India Vietnam
Philippines Eastern Europe had an arbitrage around technical skill they had a similar technical skill to Western economies but at a much lower cost that's gone now and the idea of abundance if everybody can build then what becomes different
in a world of abundance and that's where I think taste and empathy come to the surface so the first 95 % of building with AI is pretty easy and it's also pretty misleading I think Marcus that's probably what we meant by we feel like we work for the AI because they getting that last 5 % is incredibly challenging.
The number of friends, because AI is all I talk about now to my friends, number of friends I speak to about it and they share back to me, yeah I was using lovable, I was using Replit, I was using cursor, I built this, I built that.
I was like that's great, does it work? And the response is no, it doesn't work. Ten times out of ten.
For people who don't consider themselves like in it an AI building who wouldn't come to a meetup like this and it's hard it's hard to build something that works it's easy and that's reflective of where AI is today but also reflective of our immaturity and actually being proper builders but that's what I'm excited to learn and
learn to be good at so it's 4 a .m. in Lisbon that's been the case for me literally every day this week you can ask my wife so my wife is asleep sound asleep and then i've got agents hallucinating and the you know things are broken as they usually are
the ai doesn't know it's 4 a .m it doesn't have any empathy for me and what i've realized is back when i was running a company things would break all the time too but i had people to talk to about it.
And, you know, big customer gets upset, server falls down, a big, you know, feature launch doesn't work. That's part of the journey.
But then we would huddle together as a team and be like, okay, it's not working. Let's talk about this. Let's fix it.
And it felt like things were going to be okay because we weren't alone.
Building now with AI is actually quite lonely and I'm excited to see how many people here tonight and hope you you stick around and get to combat that loneliness that comes from just talking to agents all day
and all night.
So the question goes from can you build it to can you make it good with vibe coding? Because anybody can build now, but can you make it good?
So I've been giving a few talks, and a few talks in Lisbon as well, and everybody asks me, the number one question I get asked is,
okay, how do I make money building with AI? how do I distribute my product, how do I sell it, how do I monetize it, and my
answer is a disappointing answer and that's you actually start first have to learn how to build and practice.
And what that's looked like for me is disposable software. In the last few months I've got probably over a hundred repos, a hundred apps that I've spent no more than two to three hours with. So I've built it, I've tried it, I've prototyped it, and then I've literally thrown it away. I haven't looked at it ever again.
Who else here is like building software that is disposing of almost immediately? That's part of the journey.
And it's so counterintuitive, especially as we come into adulthood, where we're trying to make everything efficient. We're trying to be productive.
We don't have have time to waste but we have to do our homework and students know this in high school and universities we used to do throwaway things all the time but now we've forgotten how to do that we've we've we feel we've matured beyond that but building with AI brings us right back into doing our homework so doing
the homework is taking some of the ideas that the other presenters shared today and actually going and trying it, even if you don't know what you're going to do with it.
My wife asks me all the time, like, why did you build that? Because I built something and I go and show it to her. I'm so excited. And this is the reason.
We recently got married a few months ago. And I spent more time building AI prototypes for the wedding than planning the wedding. It's very frustrating for her. for her
but it was me it was a way for me to do do my homework
once you've done your homework and I keep doing you know keep doing homework
the next thing is to like build use AI to build something for yourself
who here has like used AI to build a workflow tool a to -do list something like a schedule and something for yourself
and then to those who had your hand up are you still using it one month later okay great Great.
That's really hard to do. It's very easy to build something. It's hard to get it to work. It's hard to actually get yourself to use it.
I've got a few, probably less than 10 % of what I've built. I'm actually still using a month later.
So my advice is before you think about even selling, productizing, sharing something you've ad -coded with somebody else, use it yourself for one month. And you're going to feel that friction.
You're going to be like, ah, it's buggy. I don't like the UX. Oh, it doesn't work this way. Oh, it's not that valuable
There's something else that's better out there But that's really important to experience before you even think about building a product Was supposed to show up a product that you're gonna charge somebody for and you know product eyes and distribute
So just start with the homework then work on it Using it yourself and then for others. I
I meet people that are watching what's happening with AI, I meet people that are resisting what's happening with AI, and people that are building. This room is obviously for the builders, hence why you're here. So the question is, what are you going to build?
I'm building a lot of things right now, but the project that I'm most excited about, it's not yet launched, it's called Avibin, and I'm very excited about solving this problem of how do we help people learn how to build a billion people today chat with AI I think in the future a billion people will build with AI and from a
learning education empowerment standpoint that is something that I feel a lot of purpose around a lot of mission around and something that I'm working on to support me and my mission to help a billion people learn how to build I'd
ask you all to take out your phone right now and tell me how you build how you you learned to build this is a survey I'd like you to take right now I vibe coded the survey this morning and it's just gonna walk you through a few questions share with me openly and anonymously if you come over here
the yeah everyone scan the sure sure oh the password so once you fill out the um the questionnaire which is just sharing with me anonymously how you're learning about building with ai you'll get a button where you can sign up for the wait list for Vibin
and if you just share your email address there completely optional then you'll be the first to know
hopefully in the next few weeks and I'll share with you an invite link
to be some of the first to test and to learn with me
it says feel free to join wait list okay
if that doesn't work and you want to stay with me I'm just going
to wrap up now this is how you can get in touch with me
and just message me on x or through my uh through my website and um just say hi and share with me what you're building and um would love to stay in touch and then i've also so i wrote this book called 2034 i actually wrote it three years ago it's still relevant because it takes place in the
future and i have a copy for all of you tonight as well so don't leave without grabbing one thank you thank you Thank you very much.