I Don't Like People

Goodbye Office Job, Take Care!

Let me tell you a story about someone who started as a musician, morphed into a self-appointed entrepreneur, became a crazy cat person, chased crypto and “AI” dreams, and learned to build software while breaking into pieces. Let me introduce me.

My descent began with a brief 9-to-5 game. The hands of the corporate structure and everything that comes with it wrapped around my neck like a python, and squeezed the life out of me. I hated it.

My three-days-a-week job felt like a slow torture. I tend to absorb and execute work quickly, which thankfully let me complete my assignments in one day. The other two days were begging to be filled, so I made an ethically questionable deal with my conscience: I'd use the empty hours to work on an escape plan while collecting a steady paycheck.

I crafted a crooked strategy. The company I worked for had this “flex desk” system - pick any desk you want, just move around periodically. I asked my boss if I can start early and got the green light. Every morning, I'd arrive first so I can claim the best hiding spots.

I sneakily arranged the windows on my large screen, disguising the escape plan document I was working on like a chameleon in its environment. If I ever end up in a traditional job working for you, be prepared: if the work is not stimulating, I will do the same. But do not worry much because either 1. I will choose not to work in such a position anymore - I learned to respect myself and others enough not to do it - or 2. If I do, I will leave soon anyway.

I plotted my escape through a startup idea. Having produced music since my teens, I knew there was no meaningful way to connect with successful artists and producers. So I found a real problem I wanted to conquer which sparked a solution: a platform connecting talent with established electronic music artists through 1-on-1 video feedback sessions. I partnered with someone I knew from an internship, and we secured seed funding from an angel investor through his family connections. Goodbye office job, take care!

With no coding skills back then, we spent a lot of money on outsourced development. As rookie dreamers, instead of testing the concept cheaply, we bombarded the dev team with many unnecessary features and user stories. Here’s the fun part: our users barely even touched the core 1-on-1 video feature which I thought to be the “thing”.

Three years in, we failed. Not from lack of validation - we generated revenue and even had a few golden months. We failed because I fucked up with the execution and decided to build a business based on assumptions. The company burned more than it earned so we had to flush it down the toilet.

Looking back at the end of these startup years, not everything sucked. A sweet tabby cat showed up in my yard. I'd always been a dog person - not that I disliked cats, I just never understood them since we always had dogs. I started feeding her and fell in love with her personality. Before I knew it, the cat distribution system assigned more stray cats to me and I became a crazy cat person.

Red Flag Factory

Despite the cats filling my life with joy, I hit rock bottom after the startup crash. That’s when my ex-startup partner brought an "opportunity" to the table. His connection, Counsel Cardboard - an attorney who appeared solid but was hollow inside - wanted us to build an esports startup from scratch, promising stability and a chance to continue our entrepreneurship. The deal: contract-based salary and equity that we never saw.

Counsel Cardboard’s communication style was a red flag factory. You know those company websites that throw buzzword smoke grenades at you? And after cutting through the haze of listed benefits you still have no clue what the company does? That was him personified. From the very beginning, my gut screamed "run" but my mind, still staggering from my failed startup, made me say yes.

Details felt like hot stovetops to him - he'd dance around them, afraid to get burned. We never productively discussed work hours or concrete expectations because he avoided formulating his thoughts and plans clearly. I figured I'd determine my own schedule as long as I delivered what I promised, which is the only way I operate well. I realize now I should have forced that discussion.

At least I had some freedom in this prison. A prison, just with a larger yard. I could walk around and try things, but this narcissist attorney detained me with the illusion of stability. After a year without progress toward profitability, my gut screamed again. But I kept quiet and tried new approaches, afraid to lose my financial security. Despite my disgust with the culture he created and the massive disconnect from my interests, I kept delivering conscientiously.

Budget Hermes Descended From “Mount Olympus”

While vegetating next to the dying esports venture, I started exploring Bitcoin on the side. I spotted what seemed like an untapped opportunity, worked it out, but hit a wall. That's when Maven, a good friend of mine, suggested reconnecting with an old music industry colleague who’d allegedly transformed into a successful entrepreneur. Let’s call him Budget Hermes - a cheap imitation of Hermes, the Greek god of thieves.

Budget Hermes frequently traveled to the US, dined at expensive restaurants, and talked a big game. But if you looked into his eyes, you wouldn’t find depth in there. He had a predator’s sense for sniffing out people with savings and liquid assets. He excelled at one thing only: making nothing look like everything. Nothing was ever his fault; he was perpetually the victim while others were idiots who just didn't understand his genius.

Always that "one last tiny step" standing between him and success - just needed a bit more money to unlock [bullshit_1] and [bullshit_2]. He was a sociopath at minimum who'd burn through someone's life savings just to maintain his facade of success. I know because that's exactly what he did to me.

Unfortunately, my then-naive self couldn’t see through the fog. When Budget Hermes descended from fake Mount Olympus, his smug face haloed in false glory, offering to work with me, I thought I'd hit the jackpot. He methodically let me peek into his “operations” - though calling that mess operations insults the word itself. Nothing operated; just hollow projects wrapped in golden paper stolen from the real Mount Olympus.

As I got more involved and invested, his promises multiplied like bacteria in a petri dish. He played me like a puppet and I began dancing to his tune. I worked my ass off, chasing impossible deals he brought to the table while he smooth-talked his way into more of my money. I'd already invested too much to simply admit to myself he played me.

Seven months of this dance, I finally pulled the emergency brake and fell on my face. I confronted Budget Hermes who flipped to offensive mode instantly. I gave him an ultimatum through my attorney: pay up in weeks.

I see myself as someone with confidence, but this silver-tongued devil had me questioning reality sometimes. I discovered his pattern - he'd pulled the same tricks on countless others, each time with a fresh story and shiny new "deal." He layered his victims using those who’d trusted him enough not to question his charade while sneakily burning through their savings.

The worst part? Budget Hermes showed zero remorse. He'd drained our savings, made us liquidate our investments without a gram of guilt. When I called him out, he started telling people I'd sabotaged his business. Classic cornered animal move: attack when exposed.

Budget Hermes didn't just earn my hatred - he derailed my life plans and wiped out my ability to build a home. I felt like a freaking loser. What a naive fool I'd been. My only consolation? Even Maven, with all his experience, fell for it. Not that it made me happy, but it proved Budget Hermes could fool mostly anyone. Despite my attorney’s best efforts, we couldn't extract a penny from him.

Painful to Admit, but Back Then, I Was a Crypto Innovation Tourist

I met Anchor, a U.S. attorney and innovator who became one of the few people I genuinely enjoyed working with. The only silver lining from the Budget Hermes disaster. I consider Anchor one of the most sharp-minded and enduring people I know.

Not long after connecting, we decided to build a Bitcoin payment platform where businesses could accept crypto while getting paid in FIAT near-instantly, shielding them from the volatility. Working on this gave me purpose beyond the doomed esports venture. I'd always wanted to learn to develop software, so I grabbed the opportunity and built the platform myself. I managed to deliver a working white-labeled solution; the front end mostly in Javascript and the back end running on Node.js.

Painful to admit, but back then, I was a crypto innovation tourist.

Don’t get me wrong, I did my fair share of research, dove deep into blockchain technology, and tried to understand the economics. But I missed the most crucial question: why would anyone spend their Bitcoin? The average hodler just holds. They don't want to spend. In an ideal world, Bitcoin would be stable and people would use it, spend it to buy stuff.

I brushed aside this fundamental question, went ahead and built the platform. Wonderful, functioning platform, performing well during demo sessions. People loved it, but there was no business we could capture so we failed.

Every Mouse Click Stabbed Like Steel Spikes Through My Fingers

While chasing crypto dreams on the side, I still had to keep Counsel Cardboard's esports venture alive. I desperately tried to inject some meaning into it. I designed an esports academy where beginners could learn from pros, a concept I borrowed from my previous startup.

As I always do, I wrote up the whole concept on paper, strategic plans, executions plans, made digestible versions of them for the team. Counsel Cardboard approved everything, then he shifted focus at the next meeting, trashing weeks of my work. This was not a one time thing.

I felt like a passenger in his Uber, watching him make random turns with no destination in mind. Each sudden swerve made me throw up my carefully written plans on my lap, which in absence of a sickness bag was there for me to clean up. After becoming nauseated of his own driving, Counsel Cardboard finally admitted that the esports company was going nowhere.

Together we shifted toward building a software development company. It made sense because an accomplished software architect was planning to launch her own company and wanted to join forces with someone. I loved the idea of escaping the esports chaos, but I was losing focus: wrapping up the esports company, launching a software development agency, and my things going on with Budget Hermes. Hell, it was a lot.

Despite the situation, my desire to master software engineering on a deeper level kept me going. We agreed on roles, responsibilities, and equity splits between the four of us - Counsel Cardboard included. Yeah, I know, I shouldn’t have entered into anything new with him. I know.

Our first project? Redesigning and rebuilding Counsel Cardboard's law practice website. It screamed trouble from day one: unclear direction and a bizarre mismatch between urgency and delay. They required rapid delivery but took weeks to review a simple sitemap. But we took it on, hoping to cover our underwhelming salaries while hunting for clients.

Counsel Cardboard triple-cast himself as client, investor, and partner. We agreed I'd split my time to maintain the esports venture, work on the website, and engage potential clients. The website project felt like torment. Every mouse click and keystroke stabbed like steel spikes through my fingers. After delivering the designs and requesting basic materials to continue, radio silence followed.

For months, our supposed investor, Counsel Cardboard, hadn't paid us. He did not operate as a real investor; he dispensed salaries whenever he felt like it, sometimes with month-long delays. He did not even care to incorporate a company for the software venture. Basically, we were his slaves.

He knew I needed proof of stable income for an upcoming mortgage, and even acted like he cared. When I asked about future plans, he praised my way of handling work and promised steady pay. I trusted this fucker. And delivered even after three months of no payment.

So we confronted him. He said he'd pay only after seeing yet another strategic plan - though he'd ignored every single one I’d ever created. I felt furious but put one together anyway. BS upon BS, the next meeting went off the rails and Counsel Cardboard said it’d be best to stop working together.

On one hand, I felt relieved that I don’t have to deal with him anymore, but on the other, my sense of stability and peace just flew out the window. I decided to send him a neutral email requesting payment for completed work. He replied that I should feel grateful he wasn't demanding money back. He then weaponized his law degree, threatening with legal action if I didn't shut up. Classic cornered move - just like Budget Hermes. These assholes use the same playbook.

Already worn from the Budget Hermes drama, I decided that I would let this go and gave up my right for the money. I questioned everything about and around myself: What do I do? What kind of person am I? What did I do wrong? Am I worthy of things? Am I intelligent? Am I capable of great things? Will I ever be successful? How did I ignore my gut feeling repeatedly? How will I find the drive to fight Budget Hermes? Why was I so naive?

As days passed, my dark thoughts morphed into better questions: What will I do if I meet such people in the future? How can I recognize them? What can I do to use this fallback as a fresh start? And at least I got rid of Counsel Cardboard.

Focus on Fucking up My Ventures

Life threw me into a new shitmaze. Up to that point, I'd been renting while working with an architect to design a modest house on my lot. Then my landlord decided to sell and listed the house on very short notice.

The house viewings became a weekly nightmare. I couldn't let the cats roam inside during viewings, so each time I packed them into carriers, and drove them to my parents'. With Budget Hermes having drained my savings and tricked me into liquidating investments, I couldn't afford building on my lot anymore. Finding another rental that would accept six indoor-outdoor cats within my tight deadline looked hopeless.

I figured I would do something crazy. I rented a 14m² (~150 sqft) shipping container office and plopped it on my lot next to an outdoor toilet. I’d promised myself that whenever I moved out, I would make sure the cats had a proper yard that kept them safe, so I built a tall fence with a cat containment system on top. The cats got their forever home, and I minimized my living expenses so I can focus on fucking up my ventures without worrying about basic survival. I became more “ramen profitable”.

I worked all day in my metal box, more isolated than ever. But at least my cats were safe with me, freely roaming in and out of the container through their cat door.

I Created Something Pretty Wild but Useless

You may think I'd learned my lesson about chasing shiny tech. But no - when bitterness and despair began to peel away, I booked another innovation tourism trip - this time to AI-land. I remember creating a list with a title of “List of problems - AI project opportunities”. You might see the pattern there. Language models seemed magical so I tried forcing solutions onto it.

Like a child chasing soap bubbles, I jumped on learning Python, studied language models from inside out, fine-tuned many, and even built a basic one myself. I came up with a concept of a distillation and training method that would allow even non-technical users to create, train and run language models locally on standard computers.

Wrote a paper about it, got it featured, built an MVP, and sparked interest from a leading tech company. They told me they would be open for further discussions to collaborate if I could validate the concept. That’s when the soap bubbles popped in my face - just like with the Bitcoin platform before.

The tech feels powerful, but what is it actually good for? What are the real use cases? After a period of intense searching and fighting off self-criticism, I figured that language models excel at two things: retrieving information from large texts and cases where flawed output and unreliability are actually beneficial. There are few of the latter, but I found one worth writing about later.

Neither these use cases nor the ones I discovered fit the MVP I'd built. So I created something pretty wild but useless - at least for now.

I grew to outright hate the “AI” hype, though I recognize the LLMs' place and still use them daily for tasks they excel at. By this point, I think I'd earned some perspective to judge: I'd built and trained language models, deployed them, and published a paper about a concept involving them.

As I ended my innovation tourism trip in AI-land, I wanted to rename “AI” to something more suitable. I came up with two options - I still don’t know which one I like more.

In this case, I’m assuming that “AI” equals mass-market LLMs since most people mean these LLMs when they say “AI”. Algorithmic Illusion perfectly captures what we're dealing with here. These systems aren't doing anything magical; they crunch probabilities from patterns they've seen before. While tech evangelists mention "AI understanding" and “thinking," we’re looking at a sophisticated autocomplete.

It doesn't understand or know. It's an illusion. Yeah, it seems magical and convincing - I fell for it too at first. But beneath the apparent comprehension, the algorithmic system under the hood is just predicting which words statistically fit best.

Or we could call it Apparent Intelligence. These models just appear to be intelligent, but in reality, they string together human-like responses, while there's no real understanding behind those tokens. Intelligence entails learning from experience and adapting - something these systems don’t do, despite what many believe. They only receive periodic, supervised training on new data; they don't “learn” independently.

Current Apparently Intelligent models lack true intelligence. AGI would deserve that label, but as far as I know that breakthrough hasn't happened. I still plan to work with Apparently Intelligent tech, but only if and where I see real value and fit. I've canceled my membership with the Innovation Tourism Travel Agency.

Ready to Make Things Better

How did I manage all this? I'm disciplined, I work fast but deeply, and clock insane hours. Someone once said "you get shit done". Seven days a week, ten hours minimum per day, for years. Probably a habit from my professional triathlete days, plus an OCD that forces me to tick every box on my to-do lists.

As a recovering innovation tourist living in a metal box, I swing like a pendulum. One moment I reach for purpose with a determined heart, the other, I surrender to the pull of emptiness. But I think I’ve found my way, thanks to the latest misfortune in my story: one of my cats outsmarted and escaped the containment system.

After nine days of devastating search, I found him at our old rental - he'd roamed back looking for his home. During those hellish days, I saw the limitations of current tools for finding lost pets. A problem found me instead of me chasing solutions. A real need I’d lived through myself.

So I started developing a Convolutional Neural Network model to power an app that’d help reunite lost pets with their families. Everything came together: my software and machine learning experience, my love for cats, and a firsthand understanding of the problem. After years of building things nobody wanted or needed, I found a purpose that made sense.

I’m still a broke person in a container. But I’m ready to make things better.