I Disappeared Into the Machine

For four months I disappeared into the abyss of the machine. And clawed my way out to tell the tale.

Crawled hand by hand from the belly of the whale of progress, and grind, of the acceleration explosion that we’re seeing all around us. I lost the sense of time as the updates and releases and competitors and alerts and notifications dinged off in the background, the stacked PR merges, the endless commits filling out my GitHub profile like a creeping vine along the fence line.

I haven’t written here in four months. I’ve been deep, way too deep, deeper than you should go without a flashlight or a cofounder or an income. Deep in the systems, prompts, agents and issues - rebuilding workflows ad infinitum into half-finished agent loops running into their efficient oblivions.

It was real work, but it’s been invisible. And it hasn’t made me the income I need to keep the lifestyle going. And sure, there have been a ton of wild adventures here in SE Asia. But adventures don’t pay the bills, at least not yet (crossing fingers for DiveJourney).

I basically overbuilt my whole system. Had a problem? System. Did something more than twice? New system. New release on some new model or agent workflow? New system to take advantage. And while that’s been insane, it also means I’ve been fucking around too much to get any real income or traction coming in.

That’s not to say I haven’t been getting work done. I’ve been more busy than ever in my life, and I’ve learned more about agent engineering, solo systems, dev systems, and the full stack everything than I ever had a chance to learn being a FE engineer all these years. Truly breathtaking, the amount you can learn when you build something these days and have an intelligence on par with the best engineers (and better) in the existence of humanity.

I’ve realized it’s so easy to spin up new projects and systems that are truly really good, and that code is super cheap - and not just code but good code - that it’s difficult to say no.

But no is what’s required to move forward.

At least until profitability. After that I’ll likely go back to just learning a ton of stuff for fun forever.

But for now, that superpower of artificial intelligence is the superpower siren calling to me, the optionality that I’ve always wanted. I have so many interests and ideas literally constantly competing throughout my life, and for the first time I feel like I have the ability and the tools to actually create them as fast as I can think of them (almost).

But harnesses don’t pay the bills and neither does systems thinking (unless you have a high paying job - and unless you choose to leave that high paying job to go and build your own product visions during the rising wave of AI - unless those actually do pay off - which I think they will).

The Reality Now

I have two products.

One is DiveJourney - which has shipped a ton and can always have more and can always be improved and is my passion project - but unfortunately doesn’t pay the bills yet. So it still matters and still will get some love, but a lot of that love will be time boxed and automated until there’s some cash in the bank.

Kitecraft - rebuilt repeatedly and shifted multiple times since November - is the real play and closer to a real product and main commercial bet, and will get the majority of focus moving forward.

More coming soon.

But it’s been disorienting, dare I say - discombobulating, checking X each day and having a brand new model, feature, product, release or something else that shakes up the entire ecosystem of developers, engineers and SaaS all at once. It’s insane.

I think we’re getting near the point of acceleration that we can no longer keep up with it. I know for sure that if I was working full-time at a company I don’t think I could keep up with it.

As of now, since I’m only working on my own products, I have time to read about what came out and try it out and compare it to what I have and the alternatives and think about what that means for the ecosystem, as far as I can tell, and what it means for me and my products and my existential crises.

Truly exciting times. And you get that endorphin rush when you use something new and it clicks - oh fuck - I and the rest of the world can use that for this.

The thing is, that’s happening at least once every two days, if not sometimes multiple times in the same day. I think back in the fall it was once every couple of weeks, and we’ve accelerated to roughly once a day at this point.

I don’t know how my friends at their full time jobs are keeping up. I’ve seen it said before: you either have the time to research and know what's going on with these things (because you’re unemployed) or the cash to pay for the tokens to use them.

I’m hoping I can get both.

Anyway I’m switching up the game and focusing on getting to profitability now and being more visible in what I’m learning. If I’m diving deep into this stuff I might as well share what’s worth digging into and help others out. All while getting to fix the great human trait of writing and thinking and not having an artificial intelligence communicate for you.

There is so much interesting going on, and I’d love to chase all of it. Thinking of ElevenLabs and audio products, creative side projects, video projects, weird little tools, all the infinite other ideas sitting in the backlog.

But that has to wait until I make the hard bank.

Or until I can justify going down those glorious learning side-quests.

For now, the game is profitability, locking in, and visible output.

Less disappearing into the machine. More bringing something back from it.