WARNING: a less technical, more personal read ahead!

I went down a rabbit hole today trying to write this newsletter.

You can see it with all the links I posted in the link share channel in the Discord.

I wanted to make this one really good and technical because we just crossed a major milestone!

We are solidly over 100 email subscribers to this newsletter, which is insane to me. I don't know why, but email feels like a much more personal form of communication than anything else we can be connected by, outside maybe a phone number.

I was halfway through writing a different article when I decided to start this one. That one was much more technical, but that may not be my lane (we’ll get to the impostor syndrome). That article was fueled more by anxiety and doubt of my technical abilities. It had me rethinking everything I had done in this project up until this point.

You see, I have created a lot of Jiu-Jitsu content prior to this project on Instagram. Over the last couple of years, I have posted a lot of video of me teaching and training Jiu Jitsu. I have gained single digit thousands of followers across multiple accounts. Not good, but also, not bad. With Jiu-Jitsu, I have a competition record (-99kg Round of 16, 2022 ADCC West Coast Trials, very hard to do) that I can point to as a credential and I am proud of.

With software, I don't have as impressive of a track record. I'm a Senior DevOps engineer with 11 years of enterprise experience, but I’m not “ex-FAANG” or work at an “AI Lab” like most of the top people in this space. And maybe that's where I should let the comparison end.

I’m realizing that I’m coming at this AI thing more from a practical "what does this mean for the average, everyday developer?" type of perspective rather than the turbo nerd who has to be correct all the time, whose entire life revolves around being on the bleeding edge of this stuff. Don't get me wrong, I want to learn from everyone, especially the turbo nerds (I’m a nerd too, don’t worry). They are, by definition, the best at this kind of thing. I truly believe whoever spends the most time on any skill has the most say in the conversation.

But I also don't take my own technical abilities that seriously. My identity isn't wrapped up in how technical, complex, and correct I can be. I run into trouble (cognitive dissonance) when I try to pretend to be someone who I am not.

I think my lane may be translating what the turbo nerds are saying on the bleeding edge into something more actionable for the average, everyday developer who doesn't have an extra part-time job's worth of time to dedicate to the stuff.

I do this best by livestreaming myself using the latest tools and techniques. I want to show the real process of learning something brand new with all the sharp edges, gotchas, and frustrations along the way. I'm a true believer in the iterative, “learn-by-doing” approach, and this is just a natural extension of it.

I think that's what I'm good at: being a crash test dummy for all these new things and sorting out what's real, what's not, what's practical, and what's a high ROI for the average developer. I think I'm firmly in the “early adopter” camp of this AI coding adoption curve, and everyone reading this is, at the minimum, early adopter to early majority.

For some reason, these 100 email subscribers mean something more to me than all of the random Jiu-Jitsu Instagram followers (outside of people I know) that I’ve picked up over the years [1]. In strictly business speak, it is more friction to put in your email for something than to hit "Follow" on a social media app, therefore, the email is higher intent from the consumer and more valuable. Also, most of your feed is algorithmic and interest-based anyway, so followers don't necessarily see everything you put out in this new social media landscape. E-mails are guaranteed to get delivered to your inbox, social media posts aren't.

I guess it's a sign that I broke through a mental block for myself that I can sound intelligent in technical content vs. sounding like I’m decently okay at Jiu-Jitsu with a non-traditional teaching style.

I guess I've always had a sort of self-imposed limit on my technical ability / interest. I hung with some “wicked smaht” people in college, and I may be the least successful of my graduating Computer Science friend group (love you guys). I saw that they are just as human as I am, but I wasn't “all in” on this software engineering thing like they were. I had this Jiu-Jitsu thing in the back of my mind that I wanted to go and compete in first.

Well, I've gone and competed in the Jiu-Jitsu thing, and I'm pretty much done competing due to a number of unintentional blows to the head over the years. There's no striking (punching and kicking) in Jiu-Jitsu explicitly, but you're going to take some accidental knees and elbows to the face as you're jockeying for access to inside position. You will also face-plant (from standing!) due to not wanting to concede and let go of a position to post with your hand vs. your head on the mat with all your body weight on it. Lessons learned.

Going from competing in Jiu-Jitsu to now focusing on competing in this AI coding content creator space thing is different, but also the same.

I started to spiral on this imposter syndrome thing when I saw people introduce themselves on the Discord. It just finally hit me the type of people who, A. Watch me fumble around with agents on live stream who then B. Decide to join my silly little Discord server anyway.

These people are tenured, high caliber professionals who are trying to catch each stream live. Who tf am I? I'm feeling a little bit like an imposter.

It almost feels like I’m a purple belt in Jiu-Jitsu (the third out of five belts) teaching an all levels class with brand new white belts and highly seasoned black belts, which I have done before. It feels like many others in the room should be teaching instead of me because they are higher ranked than me, yet here I am in charge of the class because somebody higher up believes in my ability to consistently show up and explain what I know to be true about Jiu-Jitsu.

And that’s the ticket. As long as I'm speaking from my own experience and demonstrating it on livestream, I don't think I can really go wrong or say anything terribly incorrect. I may use a tool wrong or be uninformed about something, but as long as I say things from a perspective of "this has worked or not worked for me" and not from a prescriptive stance of "I’m an expert and this is how you should do x, y, and z," I think I can just be me and be okay.

I think not being super attached to the outcome or identifying with how correct I am is an advantage because I am super open to learning about new things and I'm not too tied to any particular way of thinking. I do think there is a momentum in skill acquisition to going down a certain path and advantages that accumulate by spending a lot of time in a certain niche.

I do advocate that we pick one tool and stick with it for a decent amount of time to really learn it, but we should be ready to give up our favorite tool if something clearly better comes along. The principles that we've learned along the way will hold regardless of specific implementation of the tool. There is no such thing as “bad experience”.

I had to reconsider Gas Town this week by trying Claude Code Teams with the built-in Claude Tasks because I feel like this may be what all the “normies” will use (I consider myself to be one). Some amount of Agent orchestration will become a common denominator that will be baked into all Frontier Agents soon enough.

That was six weeks of using an early alpha tool written by one guy (and then an open source community), and now maybe I don't pick it up again until it's in a much more production-ready state because there's now a built-in, “stock” (as in OEM) agent orchestration tool in Claude that is now supported by a multi-billion-dollar company.

This type of thinking may potentially be what's drawing people to me? I'm willing to be wrong and reconsider my opinions. I'm not overly identifying in one camp or another. I do take certain people in the industry's words to heart more seriously than others, but I'm not here to fight technical religious wars like Vim vs Emacs or Tabs vs Spaces.

I'm here to just be a regular guy who's trying to surf this incoming AI coding wave and at the minimum, survive, and hopefully thrive off of it. I am high conviction on direction, but open to being wrong about what that exactly looks like week to week.

There’s a handful of peer livestreaming vibe-coding creators that I guess you could say we’re “in competition with” each other in the loosest of terms. There’s only so much space on YouTube, only so many eyeballs to absorb your content. Your content has to stand out above everyone else’s in your niche.

I would like to think that my content is more… relaxing than theirs? They have titles like "Day 123 of Vibe Coding to a Zillion Dollars", which is cool. If you're a young, ambitious, 20-something person, I totally understand the appeal of that, and more power to you. I too, was once young, scrappy, and hungry, and I didn't throw away my shot.

I guess I'm cornering the potentially 30+ crowd who are deep in their careers already, have the weight of the world on their shoulders, not a lot of time, and just want to hang out and learn this new fun thing that is completely changing how they do their jobs. People who want to learn for the sake of learning. If we make money along the way by doing our jobs better or by creating something that other people use, cool. If not, the process of learning itself and the fun we have along the way with others is the reward.

I guess my mission statement is to “be a good hang”. That's what I've gotten out of watching others on YouTube personally. The news and what's going on in the world is distressing enough. I just want to be able to escape into a one-way parasocial conversation that somebody is having about their hyper-niche interest like Star Wars or Gundam Wing. That's the use case I have for YouTube and I imagine others do as well by watching me fumble around with Agents and yapping about it.

Like I've mentioned in a previous newsletter (I freaking love citing myself), the future is going to be much more entrepreneurial and distribution is going to be key for anything that we do. By default, nobody knows that you exist. Even if you built the best app ever, you need to be able to get it in front of people, and people have to build trust in it because there's likely going to be dozens of players in whatever space you're building your app in due to the “Forest of Low-Hanging Fruit”.

Trust is built much easier with your face and name attached to it. It's a certain level of accountability. It gives people encouragement that there's a real human on the other side with a personality. Not just some AI bot that is running rampant and is trying to make money to pay for its own electricity/AWS bill or to pay humans to run physical errands for it.

I recently read this article that is 14 years old but still rings true with what I'm doing currently by expressing myself on the internet in the age of AI-generated “slop”. I produce my own slop too for short-form distribution. I'm not railing too hard on it, but rest-assured, this newsletter is all hand-(or speech to text)-written. People still want to buy from people, and if that personality leaks into your product, all the better.

I think I've been so busy trying to carve out my own path in this AI coding space, I forgot to stop and look around at what other people were doing, and that kind of tripped me up this week. I've basically just been following Gene Kim and Steve Yegge’s lead and was ignoring basically everything else. I think I needed to do that to get this running start in this particular direction.

With some encouragement from a business mentor (thanks Scott!), he suggested that I write this weekly newsletter and create content around a particular niche that I have expertise and a unique viewpoint in. Here we are at the seventh issue and we have over 100 people who are going to be reading this in their emails alone with another 240-ish on LinkedIn.

I just needed to put my head down and ignore everything else that's out there, and develop my own opinions and convictions around this AI-assisted coding thing that Kim and Yegge enshrined in a book called, "Vibe Coding." I'm finding out that it isn't the end-all be-all, or only thing to read or learn, but it is a great starter text.

I don't regret my choice at all of pursuing "learn by doing" by just focusing on what I can accomplish with the agents on livestream, and just end up yapping most of the time anyways. You can expect more of the same moving forward.

If you're here reading this, I just wanted to say, thank you. I hope you're getting something out of this, even if it's just enjoyment. Maybe most importantly, enjoyment.

I want to encourage everyone who is on the fence about doing something creative to just do it. You don't need anybody's permission. If nobody sees it, nobody sees it. Meaning, don’t worry if you're not good at first, because the algorithms won't push you, and you won't be that embarrassed. You're just expressing yourself, from your own perspective, from your own corner of the world

If you're interested in writing a newsletter in particular, here's my affiliate link. Otherwise, just go to https://www.beehiiv.com/ to get started.

[1] Not that it isn't cool to be recognized from my content by some people who come in to train at 6am at 10th Planet Las Vegas, it definitely is and I try to keep that in perspective. It's awesome to see the real-world implications for algorithms. Every “view” you get on a social media post means that a pair of eyeballs that are attached to another human (probably) saw your stuff on the other end of the internet. It also definitely works in my favor that I teach on some of the more recognizable mats on the internet, most notably due to our black belt competitor Andy Varela.

But Jiu-Jitsu is also a small world, and I've been in it for 13 years. You naturally cross-train at other gyms when you travel, and lots of people come through Vegas. Now I have friends all across the country and the world. Even if I didn't train in Vegas, I would probably get to know a decent percentage of the Jiu-Jitsu world because there's “only” 6 million people who practice Jiu-Jitsu on the planet and ~1 million of those people live in the U.S.

Comparing this to an estimated 4.4 million software developers in the US, I am at minimum 4x more likely to meet a software person than a jiu-jitsu person. For jobs, this means 4x the number of competitors to "win” a roll (c wut I did thar?) at a company. There's just a simply a different level of technical skill possible with the sheer number of people who are practicing in a given activity.

The wider the base of people practicing the skill, the taller the possible absolute height of the pyramid. It's like math or something. Same thing happens in other highly in-demand professions. Think of a major league sports player and then think how many hundreds to thousands of kids that they played with growing up who did not make the pros while they did.