April 7, 2025. MIKE-AI.

Once I promised I’d drop more info on MIKE, this project I’ve been obsessed with at Seraphim. I’m keeping some details on the low for now, not to be shady or anything, but some stuff just isn’t ready to hit the spotlight. You know how the internet is (once it’s out there, it’s out there forever).

MIKE kicked off for a couple of weeks now. MIKE isn’t like ChatGPT or Claude or those other AIs. Those are like that janky Swiss Army knife you got for camping, tons of tools, but none of them really hit. MIKE’s more like a surgeon’s scalpel. It’s got one job, and it does it with crazy precision. Specialized, not trying to be some over-the-top all-in-one deal.

Building it has been weird as hell, no lie. I’ve been like no sleep for days straight with random solutions to problems I couldn’t crack all day. 

I started keeping my notebook by my bed ‘cause those ideas? They vanish by morning if I don’t jot them down. Half the time, my handwriting’s so trash I can’t even read it the next day, but somehow those sloppy notes turn into legit breakthroughs. 

Honeslty It’s chaotic, but it’s working (so okay 👍🏻).

The wildest part isn’t even the tech, though. It’s these big, messy philosophical questions that keep popping up. When you’re coding an algorithm to figure out strategic stuff, you’re basically turning values into math. 

Like, what’s “good” supposed to mean? Who decides that? How do you weigh all the trade-offs? These aren’t just coding headaches, they’re human problems with no clean answers. 

Some days, I feel like I’m making up a whole new language to turn vibes into code (code vibings you know). Other days, it’s like the dots were already there, and I’m just connecting them. It’s trippy, like those quantum states I won’t shut up about (possibilities crashing into actual code).

I’ve messed up plenty, too. The first version? Stiff as hell. It crushed stuff already knew, but throw something new its way, and it was game over, a completely total crash vibes. Not cool for something that’s supposed to handle the unpredictable. We’re on version 5 now, tweaking it to adapt and reason from scratch instead of just riding patterns.

Progress is all over the place. there's no any correct numbers to tell the whole story. Development’s not some smooth road, it’s more like random leaps between “why is this broken” slumps. I’ll bang my head on the same problem for days, then bam, solve three huge things in one night. 

It’s a rollercoaster, fam.

I also tried that “build in public” thing, sharing updates, chasing feedback, hyping the vision. Truth is, most people don’t care about your ideas or plans. They want proof, users, cash, results. 

At first, that stung, but now it’s kind of a relief. I’m not sweating if my choices are “perfect” anymore. I’m just going with what I know, taking some wild swings, and seeing what lands. 

I don’t trip over how other founders might play it, as long as I’m real to the vision, the work stays solid, and the outside chatter just fades.

Right now, I dialed in on making MIKE killer at finding strategic leverage points, those clutch spots where tiny moves flip the whole game. That’s where it could outshine humans, not ‘cause it’s “smarter,” but ‘cause it can crunch a million variables without getting messy with biases or emotions.

Big lesson from this grind: building something dope doesn’t make it a business. People need to see real value or at least something worth paying for. And some problems? They’re not even fixable, they’re just part of the chaos you sidestep or roll with.

So, am I even doing this right with MIKE? No clue, fam. Some nights, I wonder if I’m just a mad man holed up in my room, coding while the world keeps spinning. But then those breakthrough moments hit, when it all clicks, and I’m like, “Wait, this might actually slap.” 

Still, I’m trusting the grind, ready to switch it up if I need to, and just keep pushing. 

That’s the game, right?

For now, I’m staying low-key on MIKE ‘til at least May 10 (let's see how much i can deliver with MIKE). Less yapping, more doing. 

Catch you later, fam!