Am I Alone transcript

September 1st, Fable 5:1 drops. September 2nd, Gemini 3:8 slash drops. Of course, what it lacks in general intelligence, it makes up for with Mr. Beast. And I can't wait to find out which developer is going to make it the longest. 3 hours later after Gemini has been released, Muse Spark 1:3 gets released. And Muse Spark is near the top of the artificial index intelligence list. I was not expecting that from Meta. And then Astral.

Astral is released and all the models go down. Chat GPT 404s and everybody says that AGI is here, including Greg Brockman. Then what happens next? The inevitable onslaught of game slop. We have game slop and game slop and game slop game slop game slop and game slop. All of it showing that Astra is in fact the greatest model to ever have been dropped. And guess what? You can't use it. And with all these models coming out and and you not even being able to use Astra, the worst part about it all is that Claude gooning on main, okay?

This is a mistake. AI is in fact a mistake. But I'm tired, boss. I am I'm tired, boss. I don't know. I'm getting new model fatigue. Every time I see these new releases, I just keep getting less and less excited. It kind of seems like since Fable 5, I haven't really felt, I guess, the level of the same excitement. Fable 5 did kind of feel a little bit more special, like, okay, I could really code with this. Grok 4:5 and then Grok 4:6, uh the the speed at which things happen, I felt like, okay, I'm really I really like this.

This is much faster. I don't have to have like a 50-minute wait. I actually have like a small wait. And it kind of felt a little bit more special. So, I was generally excited to see these things kind of drop. But this week, when all these models dropped, I didn't really feel like some amazing, like, oh my gosh, it's the next coming. Like, I just didn't feel that. And I'm going to analyze myself here, and it's going to be a little bit vulnerable, okay?

Part of me wonders the following. What was the purpose of coding? Well, there was the learning aspect, doing recreational coding so that I could become a better coder, so that I could build better better and bigger software, but the other half was to make developers happy. Like, that's what I did at Netflix, and that's what I really enjoyed was like, can I build the right tool to make stakeholders or other developers really happy to make their life better?

And that's something I really, really enjoyed. And what you see behind me is me doing it again, but I'm slopping it together, and part of me just gets so excited about that. It just makes me so so happy to actually like make someone's life feel a little bit better, like they don't have to do as much, or that their tools work really, really well together. And so, me building this Omachi automation, it actually makes me happy.

I feel like I'm actually doing something. I'm trying to fix a real, real problem. And yeah, I slopped this together. Yes, all the logs and all the actions are stored in PlanetScale. It's done with a client-server architecture, so I can use cloud agents with Cursor. The tickets on Linear are keeping track of everything. I'm kicking off actions from Linear. And yes, Sentry's actually measuring out all the little details, so I can go back and review like, why did the session take so long?

What was like making the model act slow? And it's an incredible system, and I can't believe I was able to prompt my way to it. Yet, when I see all these new models, I'm not feeling something like, oh my gosh, I have so much more intelligence. And I have to say, I think I understand why this is happening. The reality to the situation is that I've had to let part of myself die to be able to run this fast. And what I'm saying with that is that there was something I enjoyed so much.

I enjoyed putting on the headset and just programming for 6, 8, 10, 12, 16 hours. I could just go so long and enjoy just love it. I remember just jamming to music and just crushing the keys, and there was something so special to that. But, it's impossible, at least it feels to me, to really look at that as a viable solution today, because when I look at the speed in which I could create the system, which would have would have taken me a couple weeks to really tune and make sure it's correct along with just kind of the testing I had to go through and all that.

And it'd actually be very hard to build what I built and it would be some really strong heads down time and now I was able to accomplish that in approximately two days. Like I have to admit, yes, this is fantastic which means that the thing that I really was good at which was neovim programming is less valuable. I know that it's hard to admit and I think that that's part of the reason why I'm just like the new models.

I I just don't feel that excitement and I I I I I think I kind of realize why. I was already able to build everything I was really excited about with Fable, with Grok 46. And so when these new models come out, yeah, I'm probably going to be able to improve stuff, spot some bugs, kind of prevent some stupid behavior and it's going to be awesome. But at the same time, I'm still going to have the real problem I have which is how do you make someone's life better with the code, which is really how do I make it better with the project.

And so it really doesn't absolve me of the hardest part of programming, which is making something awesome. And so when I see all these slop games that you see, like there's a reason why I feel upset and it's because the actual hard part about making a game isn't putting together a demo. It's making something that's repeatable and fun and when I look at that I go, "Okay, yeah, that took you some money and then waiting for a prompt to finish, but the actual fun part, it's not there.

There's nothing fun about what you built." And really that's a reflection of me, which is that given infinite resources, can I build something fun? Can I build something good? And yes, I would like to say I'm building something good right now. As you can see the full Omachi experience all the way to the lock menu and then resigning back in. Like that's awesome. And the fact that I can actually view a session happening between a client and a server by following along what's happening in the database and having the actual images show up in the terminal is an insane experience.

And yet, here I am. Not feeling the excitement. I don't feel the same things that Twitter does. I don't I I'm not caught up, I guess, in the LLM new release psychosis. And I think part of me is just I I am that way because the thing that I enjoyed so much about programming was a selfish activity. I enjoyed code masturbation. I'm going to say it. I enjoyed the process of typing and building the things. I enjoyed the process of thinking through everything.

And really, you have to admit, code masturbation's not a bad term for it because at the end of the day, I've said this a thousand times. I said this in 2000 and 2020. And I've said it in 2022, and I'll say it again today. Nobody actually cares about the inside of what the product looks like. They really care, is it good? Is it smooth? Does it feel well? Is it a laggy piece of crap, or is it actually a something that is really smooth and amazing?

And so, honestly, I know I have new model fatigue, but I'm trying to turn something around inside of me, which is I can't believe that we live in a world where it has never been easier to build something that you've always kind of wanted to build. The weekend project you can take on an infinity scope, comparatively, and build something, and that is going to feel awesome. And so, this is me attempting to change something about myself.

And this is kind of the video of me explaining to to change something about myself. I've always wanted to make a dev channel that's very uplifting and positive. And always constantly being upset at the state of the world is just like that's not positive. That's not good. Like, yeah, it's hilarious to make fun of Google. I'll continue to make fun of Google, Microsoft, all day long, and Sam Altman, and Dario, and all that.

We'll still keep doing that. But, I don't want to live a life where I'm upset about technology. I want to look at it as like all the best parts. Like, what a wild and magical world we live in, where we can make something better. So, yes, I guess reframing it that way and becoming excited about the potential, it makes me really happy. That means I can take this project, which is going to save many, many people many, many hours, and ultimately, it's going to hopefully find a bunch of super cool security issues and be able to flag them and fix them all so other people don't actually get their system hurt like that.

You know what? At the end of the day, that's amazing. That's what the entire goal was whether I hand-coded it or not and I'm just going to focus on that and try to really enjoy the process along the way. Try to find a way to learn in this new era. How do I take the skills I already have and how do I turbocharge them and make them awesome? And so that's that's me just kind of talking to you guys through everything and I know there's going to be a lot of people that are pissed we lost my I know I'm going to see in the comments a bunch of people saying we lost him.

He's gone. The last one you weren't supposed to join the Sith. You were supposed to defeat him. But hey, at the end of the day I'm a techno I'm a I'm a tech a technology enthusiast. I want to see the good and I want to make tools that make developers' lives better. And if this can make me make the best tool that I would just not dream of making because it would just take entirely too much time to do it by hand, then guess what?

I'm going to make that. Honestly, this video was kind of a a therapy session for me. Yeah. I'm happy about this video. A gem.