Hello,
Some of you may remember me, some not, I've been out of TFLF for a year now.
I've learned much more from execution and working 100-120 hours a week for all the 4 years than from any thread from this forum, and I had to learn on my own mistakes and advise from the few around me. I wish that I read this thread 4 years ago.
I'm 21 and down financially, so if you are older, or are well off, the thread doesn't apply to you. However, if you are a young entrepreneur starting a company fresh out of college or you know somebody who is, that's where I was in 4 years ago.
I write out 13 learnings/retrospectives - mistakes that were made that have cumulatively cost me 4 years with very little upside. After the retrospectives, I write out my life story for the last 4 years to show how your life may play out if you decide to start a company without experience. At the end, there are open decisions of me trying to get back onto the average, stable but well-paid life.
In part being influenced by MJ's books, I believed that you can just learn your way out to every problem. Probably yes, but this is like scaling not one, but 10 mountains at the same time, each mountain being a skill + your business. Some preach that you can't just learn it. You won't, see #9.
#2, #3, #4 just go deeper on it.
Retrospective #2: starting a technology company without experience of developing with that particular technology is infeasible. So is without sales expertise.
Retrospective #3: You can't learn some things online
I never understood how to do good UI design; I still haven't found any course over the years. Things like complex AI research are also in the same area.
Retrospective #4: You won't make money in business if you are not an expert. Become an expert first.
I personally never seen a business owner who started a business and then they just succeeded in a new field. Johnny boy was a car salesperson for years. I bet that helped. if you haven't 3 YoE - you probably shouldn't even try starting a business. Nor you are in a financial position to.
At least that's what I'm coming to realize now.
As YCombinator said: "if you aren't hireable by one of our top startups, we won't fund you" (they won't because you won't succeed)
#5: People who are working (knowledge) jobs are NOT stupid despite self-improvement fellas telling you they are.
Where I work, and in my college, people are more disciplined than most who come (came, before the 5$/month filter) on this forum. And oftentimes, they are more knowledgeable too. I personally don't understand why in self-improvement, where I've discovered about entrepreneuship, being an employee was treated derogatory. I've stayed away from employment for years and pushed on business without becoming an expert first and it hurt me bad. I wish I didn't.
#6: getting and working an expert job for extended time will teach you much more than years of being a failing solopreneur
Fact; it's hard to get a job, and it's even more complex to get paid consistently later. You need to prepare months ahead, and you need to work 40-60 hours a week to stay. Maybe more. Because somebody else will take your place. And you will panic because if you don't deliver by a deadline, you won't be paid.
But you know what? You may not be paid a lot in the beginning, but you are being paid to learn from your own mistakes too. No amount of trial and error in a business has told me as much as a job did.
#7: Switching projects is deadly. And there is pressure to switch.
Building projects is like building a rocket - it may fly, but it certainly won't if you run out of steam before you can complete it.
Common lists of problems pushing you to switch a project:
- You are out of money to continue, so you postpone it and never come back,
- You don't know how to ship X,
- You are getting issues from elsewhere (personal or otherwise),
- You realize that the project shouldn't have been started in the first place.
And you quit midway. But the rocket that you've built will not fly. It will rot, and soon enough, you'll need to build a new one.
"rockets" may be skills and domain knowledge too.
#8: Gains from "outworking others" are not real if you are switching projects:
This. People who are on the top of their fields - that I look up to - have compounded their work - and I mean real, verifiable work like:
- publishing papers if you want to do a PhD,
- having income,
- having company names on your resume if you want to have a job,
- going out there and trying things if you want to start a company,
- selling to others if you want a sales position.
If you switch 10 projects in 3 years like I did AND you aren't an expert in your field, you will achieve nothing but wasting months of your time (and in case of software, stressing servers of Github).
#9: When you start a business, you will stop learning "hard skills".
Tying into #1, #2, #3, #4:
You won't or you will spend incredible quantity of time learning hard skills and domain knowledge before you can turn a profit. If you want to start a software company (picking an example from my day job) doing analytics with AI for ecommerce businesses, and you didn't have experience in both software AND analytics AND ecommerice, you won't complete it. Where I work, for example, the only reason we do have it is that:
1. Our founder had already owned a ecom business,
2. He hired great talent for his work, and I'm talking of our technical leadership who work 80 hours a week for a not very high wage; maybe equity.
What's worse, is that if you try to learn either or both of the two, by the time you will in fact finish learning, the marketplace will change and you'll need to continue learning more; and when it'll be time to serve your product, your knowledge will still be shallow.
Citing #4 - you won't make money if you are not a domain expert. In general.
#10: You can't fallback to a well-paid job that you want to quit to if you fail your business.
I heard advice that you'll have higher salary if you start a startup. From multiple places - YCombinator, and others. This is a lie. In fact, you'll be treated as having a career break, because it is in the eyes of the employers. So that job at Google that you want to take after starting a company? no no.
#11: STEM degrees pay (unlike what MJ says.).
I don't know one would advise against colleges - probably because in the US they are expensive. It appears to me that people who advise against degrees didn't pick the right ones.
I never had somebody advise me against a degree in-person, though sure I've heard lots from the internet (including from books many of us have read)
yes, it takes learning, but you are far better of having a proper salary right after college and subsequent years - and in STEM the salaries are good. Having a good salary allows you to do business, and being an expert that others want to work with. If you are a plumber, you have the same pay as somebody from college, but don't have the privilege of #2. And you don't have the expertise necessary for starting anything but a plumbing business because #9.
#12: You can't start a company if you don't have capital.
8/12 of my business attempts over the last 4 years stopped because I couldn't continue financially. Get a paid skill before you start a system.
#13: Wealth is built incrementally, and I've never seen anybody do the first few steps from a tech business.
If you are there like me building a company after college, you will not be able to hire others. And that's preventing you from getting skills you don't have or buying technology that you need. For example, I couldn't buy enough AI compute to finetune AI models properly, because those needs hundreds of dollars. And I don't have hundreds of dollars lying around. Moreover, I don't have others to hire to streamline the process.
citing MJ's video, I subscribe under every his word "From Flat-Broke to Filthy Rich: How to Get Started Toward Wealth When You Have No Money & No Skills" :
View: https://www.youtube.com/watch?v=r_58zPgArb4
That's it from the learning that I had.
The rest of what I learned were only technical skills, really. Nothing that most people will care of on this forum, as far as I understand.
But I did try in that college. By the end of 2021/2022 I was the top of the class because I was sick of bad grades and for the first time, decided to learn.
Just 5 months later after I started learning software engineering, I got a job paying 8$/hour - well above average for Slovakia, where the average was 4$/hr.
But I felt pull towards business and disregarded that job. I planned out my business when I was supposed to be working. moreover, I didn't even understand what the product we were developing was about. I had a 2-hour zoom meeting with an engineer, but I was too green to understand that that was enough to start working; because I didn't know how to work.. I was able to write and read code, and did ship two or three PRs over the course of the three months I was on the job, but that was it - i didn't even understand what problem the project was solving. I understood it just a few days before I was fired.
I did code my software project meanwhile. I actively spoke to users (every day, found conversations through Reddit). I wrote the software - it was difficult and I had to learn multiple languages, but I did. and I learned a slight bit of web design. The truth is, I wasn't an expert not at the software not at the domain I was building it in; I had no idea what "marketing" meant or how to even sell something to somebody - I never had any experience selling, or working anything.
So I just developed it day in and day out. The result was so-so because - I wasn't an expert.
I've listened to YCombinator videos and listened the promises of them telling that you can become a billionare in the field and that it's better than a job. Yeah. Except sometime until 2026, the job I was fired from was the last income I've made from software for 5 years.
I don't have much to inject from what is already said, but this sets the starting point.
(retrospectives #1-4, #7, #9, apply)
I was on family money support.
(retrospectives #1-4, #9, #12)
Decided to stop a long-running project of FullCook (on which I worked for 10 months). Decided to do an AI that will change the world (quite literally). The idea: make an AI that can be a hardware engineerz. It didn't exist, I've double-checked. I spent 4 months doing an environment in which the model could be trained. My Bachelors degree is over.
Moved to Ireland. In Ireland, lived for welfare because we were given it. Shame, but I spent a year on a welfare doing this project. I think that had I not stopped on FullCook, I could have continued it as a project and it could be a meaningful business by the end of 2024.
(retrospectives #1-4, #7-9, #12)
First half of 2025:
I've Learned AI and Python; software engineering as a whole. For real this time, from the projects and many models that I trained while learning.
After learning machine learning, software engineering, switch projects. Started more startups, spending a month on each. A project in robotic assembly which I never finished because again - I had no money and my knowledge of AI was still shallow. A project that was meant to accelerate training of AI/LLM models. Also didn't finish or publish it because I had no money.
Second half of 2025:
Welfare money ran out mid-2025. I moved to Dublin: I was certain that I would find a job and there are more money in Dublin, even in hospitality, but I couldn't. - AI - - AI entrepoorneur's resume for hospitality jobs
Money has almost run out. Started window cleaning again! Doing 200$/day - WOW!!! most money I've made in my life. Paying my own bills.
My family pushed me get a degree, which I enroll into. Tentatively, I get a Master's degree for 9000$ in tuition paid in installments over the next year. It started in January 2026 and lasted one year. It appears as a OK college.
Continue starting 3 more startups in all my free time.
I started daydreaming of becoming a software engineer at Google or in any appropriate company - after all software engineer are making lots of money - in Google they said 140-300k/year depending on seniority + equity.
I'm hoping for the best, and I'm more software as startups until November
October 2025 and I'm cleaning windows.
At november and December, I've committed to window cleaning full-time. Wake up at 8am, start window cleaning by 10am, finish at 8pm. Watch youtube videos which talk about it. Made something like 5k/mo working 6 days a week, hired an employee, ran ads, bought new gear - lost half of it.
(Retrospectives #5, #9 and #13)
I started 2026 by paying off my college tuition. Saving money from window cleaning, I've paid off everything that I had for two months for college. I resumed did window cleaning from January 5th to 15th, but suddently, I was making 2x below minimum wage - 5-6 EUR an hour. Nobody wanted their windows cleaned in January! For days straight, I came but with nothing in my pockets, soaked in rain. By February, I was out of savings.
I've paid 120$ for a Jobber CRM in November, and I hated every bit of it. it took me 4 minutes to enter details of a client! so having free time and not doing window cleaning, I decided to make a project doing a CRM which would be much faster to use than jobber was. The idea: text all messages into Whatsapp instead of a CRM. You could type something like "add client john window cleaning, 50$, Millstreet 5" and it would log a job just as a CRM would. I made a UI too.
I've stopped after two weeks of intense full-time developement because we needed to do a project in school and I decided to revive my idea from 2024. I also wasn't sure anybody would actually buy the product. Nobody ever knew about the project until you reading this.
February:
We had classes of AI. Because I've spent the whole 2024/2025 learning machine learning I was fairly competent and well, I didn't learn much. I always made it a point to show up and comment and be active in the class.
We had a project due by March. So I decided to revive the engineering AI startup, trying to make a publishable research (a paper) about it, on how to evaluate and how to improve capabilities of LLMs on it. It took me whole 2 months of in and out of developing it, and it barely worked at the end. Actually, it did work as intended, but LLMs were in fact incapable of solving the problems - not a single problem was solved within a time budget. Mind you, I had no money to invest any further, so the project was shutdown.
I've also gotten an offer from an engineer to join a startup, where CEO offered free hours and remote job. I hesitated - will I come back to a job after success in window cleaning? I've spoken with a CEO and the contract was a unpaid internship.
At that point, I've written code for two years with 0 income from it, so I accepted it.
I've reengineered their entire application - their backend, UI, added a pile of useful features, and left it for their review. Incrementally allowed them to code review it. they responded after 3 days after each of my message. At some point, I realized the startup was in fact running for two years and had engineers shipping features very occasionally.
I was invited to be paid for my work - 1.5k$ for what for me was 10 days of work. I said that - maybe I should become their technical co-founder? they didn't say yes for no reason in particular. So it was a no.
Restarted window cleaning on March 15th. Consistently running out of savings though because college + own startup + window cleaning + working for the startup - the window cleaning didn't fit in. Needed money support for tuition for my parents. Always paid my own rent though.
March:
Money-wise, march was the same. Window cleaning pays bills, pays tuition, pays food. I am developing the AI paper.
Two people startupschool.org have contacted me during March. Two potential co-founders. One suggested doing a software project together and one asked if I was interested to contribute to his project.
The former was a business owner with a track record of selling websites and marketing to business owners. Said he and his co-founder were scaling up a marketing agency to up to 30k profit/mo in one month. then they part ways with his cofounder, because his cofounder wasn't interested in software. He looked on me to start a software company with him. I was interested, so I did accept it. We discussed it. I went to different city to speak to potential clients. The clients informed us that the idea was executed many times already; in particular, his idea was software for quotation of houses. So I have reinforced that because there are literally >20 companies doing the same, we shouldn't do it too.
April:
I've gotten an offer from the "contribute to his project" startup founder whom I've met at startupschool.org. He turned out to be Russian. He offered me to join his company (Nyle AI Unmatched Tool for Amazon, Shopify and TikTok Shop Sellers) for 2k$/mo. He said: "We need to ship features very intensely for 2-3 months and then I'll be able to raise wages because features will be built". I've hesitated, asked for advice from software engineering roommates who did 90k$/yr (fine by Irish standards) and was advised to take the job. I did take the job. This was my first real job.
I passed an interview due to my machine learning skills; where I did the system they asked for in a day.
The founder has promised strong salary gains if I performed well.
The codebase had no documentation whatsoever and I had a contact with only the lead engineer and a non-technical founder.
I immediately jumped into vibecoding the features they wanted or needed, talking directly to the non-technical founder. Merged 10 PRs after 14 days. Felt good about myself.
Retrospective #1: wow: real jobs are real.
Retrospective #2: Wow, did I get paid money to write software for the first time in two years?
Retrospective #3: I still wasn't explained what the project was about properly, and was outside of the team for 14 days,
----
On starting own software company side, I paused it or contributed to it for 10 hours a week because I had a job and if I wanted to be somebody like that 90k/year roommate (and I wanted to), I knew I needed to push my job hard).
May:
It turned out, my new code broke a few things and a lead engineer had to fix some of it. It turned out, we have had >1m LoC in the codebase. There was real history of development. Unfortunately, it was not at all documented. It was just a pile of code and a repository from my point of view. which it was - kept in heads of two or three backend engineers.
I got integrated into the team. Turns out they had a proper workflow and proper everything, but the founder was not technical enough to know or care, much less to tell about it. the lead engineer was mute too. I couldn't have known that the issues I caused could be issues at all, but a particular external API we were working with had underlying issues too, and I have unknowingly tripped them. Had to undo much of my first months' development efforts.
after the first few weeks of being integrated with other engineers, I have for the first time shipped something useful to somebody. for real. Took rework, but I did.
Now was the time to be proud of myself.
And get paid.
I started noticing a pattern, though:
- I was working 7 days a week, 6am to 9pm.
- So did our lead engineer,
- Everyone on our team is online and working 60-80 hours a week.
- For 2-4k$ a month as I learned later.
- Also, the company founder we were with has pressured us hard at all times on shipping even faster. His derogated everyone's work in front of everyone on every daily meeting. What's happening? it was fine during 1-2 months - I closed eyes, but this was consistent.
---
starting a company side:
We paused for two weeks after I explained to the cofounder that the idea was in fact dead before starting. After two weeks, he reached out and pitched something else - to start a company building websites powered by software. Basically, that's what I did. He asked to collab with me on the repo, and I granted access (he was a better designer than me; I'm a systems engineer, not a designer) and after 3-4 days of prototyping, I shipped the first bit of software for him. It looked not the best, but it was sufficiently functional.
Then he turned off his messenger for 4 days,
afterwards, I start calling him up through any messenger I can... and get an answer:
He took my code and ran away.
What?
June:
Job-wise, June went by the same - working 7 days a week all available hours.
I was told to start a project which sounded fairly simple and I executed the first iteration of in a week or so. Then requirements changed. I needed to rework the whole thing. It took me whole june and more to rework it.
---
I stopped window cleaning. While the cofounder has ran away, I start making on another project, a window cleaning machine which I've detailed a while ago on the forum:
I've built up a simulator environment (much higher fidelity than this mockup too, where I have only experimented with pole movements) and while I wouldn't upload the whole video here, the robot could drive in a realistic environment and clean windows.
At some point the co-founder came back. I said that I'm willing to continue - just let me re-engineer a few things because I was in a rush and the first codebase was sloppy.
We agree on the following: I keep the codebase outside of his access.
I dropped the window cleaning project and was booked up again. with job + startup.
---
July:
Job: I'm stuck reworking the thing for a whole month because the requirements have changed. I didn't deliver anything meaningful for a month. I'm starting to panic. I worked for 7 days a week and got paid 2000$/mo. This is minimum wage in Ireland. I'm out of savings because I need to pay tuition too. I needed to restart cleaning windows on Saturdays.
I devoted the whole July to the job.
August:
Job: the whole project which took 1 week to do by old requirements now sprawled into 80K LoC because they asked for new features. It became unworkable. At august 1st I needed to delete all code because new features that were asked in June needed a completely different approach that rendered the first unviable.
At august 10th I'm asking "so where's my minimum wage which should've arrived at 1st, boss?". To which I'm replied "where's user features? where can I see your work?"
And I couldn't answer. I needed to rework the whole project because they hadn't told me all requirements from the start.
At august 20th I'm still not paid for July.
I had a planned vacation that was booked for my family for two weeks (August 24 to September 6th), so we flew off to the ocean. I couldn't finish reworking the major feature by that time, so I said I'm going on the vacation and will finish it afterwards.
Startup:
Since i'm on a vacation, I devoted 2 weeks to the startup. Literally polishing every small bit so that no part of the application will be slop, which would foregone the need of rework. Polishing every small part before the implementation. See: the last project was stalling because the codebase has become such a nightmare to work with due to rushed implementation that it become outright unworkable. I couldn't comprehend - or could, but with real problems - on what was happening in the codebase. It wasn't planned out, it was LLMs doing decisions and then getting tripped up by them and reworking everything.
This time I decided to do it differently. I spent 2 weeks writing 90 markdown files total outlining every small bit per feature so that nothing, nothing can go wrong when I implement.
My vacation lasted for two weeks; which I spend specifying how it will be built. By the way, this pays: you can reverse decisions in markdown 10x faster than you can in code. Better yet, you'll actually comprehend them. You can ask a model along the lines of "are there any issues" and it'll come back with an answer in 2 minutes, not in 5 or 7. You can make edits 10x faster. but your app isn't built.
September:
On Friday, I'm getting this message:
I reply that "I didn't even start coding yet". The application is very well specified; I removed all technical debt before starting. Minor bugs remain, which I proceed to patch in doc-only mode. By Sunday, I realize while the rest of the application (authentication, user management, other functionality) is very well-specified, the website editing itself needs some rework.
at monday, I'm getting a message:
What? I've just spent two weeks and we are days away from launching the product - since the application is implementation ready from a spec; I'm confident there is very little to be left to do.
I get on a call, and he says that I must wait until we meet. And that we must meet on Friday.
Two days later:
Well, I wait. I am only a technical cofounder in this business and effectively have no control whether the business is moving or not moving.
---
Job: I still wasn't paid. Will maybe be paid when I finish the project.
Our founder has held salaries for me and another engineer. I speak to another engineer, and he says he works 250-300 hours a month per his estimate for $3000. He was also, told that he would get a raise based on performance. which he never did. In fact, he had offers from better places, but our founder convinced him that "a few months and we'll get rich, and you will have high salary".
I need to continue doing window cleaning to pay rent (I'm already past due because I was on a vacation for 2 weeks). I do.
At Friday (today), I'm getting a message from the co-founder that we move over to Monday, so we postpone the meeting.
So what's happening?
---
Meanwhile, my family pushes me to get a PhD for two reasons; safety and brains. I think this is worthwhile. Given that currently I'm priced at 2k$/mo as a software engineer - and no other job offers, this appears like a sane thing to do. After all, I wanted to work in places like Google Deepmind or OpenAI or Anthropic.
I hope the story will serve useful for others - my goal was to have others to not make the mistakes that I did. For those above me, if you will ever have influence over others who want to do the mistakes that I did, please refrain them from this, for these are mistakes that cost real % of human lives.
2. Should I switch to being a software freelancer instead of looking for jobs? I've heard some have success this way.
3. This is a wrong audience to ask, but is PhD a right choice? for AI research, presumably yes. It may also open doors to any country in the world. Best of all, if I'm a student, I'm not getting departed if they cancel all our visas, and that's worth more than anything - that's an existential question at this point (that's why my family is pushing me anyway).
- If I want to apply to a PhD I need to publish research. And if I need to publish research I need to work on research. And that takes away time from job, freelancing or windows.
4. Should I stop doing any business for the next 5 years and commit to increasing my income through other means any means necessary? this appears like the right choice, given my own findings. then, PhD is conflicting with it.
5. I've already invested a month of work from the aforementioned co-founder. Do I continue? Sunken cost fallacy, but not having control over that what I need to spend weeks on is damn scary. We have shared understanding. But he says "stop", it stops. It happened twice already. but another dead project?
6. I also need to start saving up for a house in Ireland, which is well above 200k even in poorer regions. Forget about CENTS until I have a house and good passport?
I don't think I have the capacity to pick anything but a) finding a better job, b) doing a PhD, c) doing freelancing (either of, given how time-consuming all 4 are) d) doing window cleaning further to pay for tax. Startups appear out because I'm booked up until the tax is due. And maybe I should stop trying in the first place. Also, I need to prepare for the next January and February, for I will have no income if I won't have savings.
Some of you may remember me, some not, I've been out of TFLF for a year now.
I've learned much more from execution and working 100-120 hours a week for all the 4 years than from any thread from this forum, and I had to learn on my own mistakes and advise from the few around me. I wish that I read this thread 4 years ago.
I'm 21 and down financially, so if you are older, or are well off, the thread doesn't apply to you. However, if you are a young entrepreneur starting a company fresh out of college or you know somebody who is, that's where I was in 4 years ago.
I write out 13 learnings/retrospectives - mistakes that were made that have cumulatively cost me 4 years with very little upside. After the retrospectives, I write out my life story for the last 4 years to show how your life may play out if you decide to start a company without experience. At the end, there are open decisions of me trying to get back onto the average, stable but well-paid life.
Retrospectives and learning along the way:
Retrospective #1: starting a startup will not make you an expert in that field, unlike how some others will tell you.In part being influenced by MJ's books, I believed that you can just learn your way out to every problem. Probably yes, but this is like scaling not one, but 10 mountains at the same time, each mountain being a skill + your business. Some preach that you can't just learn it. You won't, see #9.
#2, #3, #4 just go deeper on it.
Retrospective #2: starting a technology company without experience of developing with that particular technology is infeasible. So is without sales expertise.
Retrospective #3: You can't learn some things online
I never understood how to do good UI design; I still haven't found any course over the years. Things like complex AI research are also in the same area.
Retrospective #4: You won't make money in business if you are not an expert. Become an expert first.
I personally never seen a business owner who started a business and then they just succeeded in a new field. Johnny boy was a car salesperson for years. I bet that helped. if you haven't 3 YoE - you probably shouldn't even try starting a business. Nor you are in a financial position to.
At least that's what I'm coming to realize now.
As YCombinator said: "if you aren't hireable by one of our top startups, we won't fund you" (they won't because you won't succeed)
#5: People who are working (knowledge) jobs are NOT stupid despite self-improvement fellas telling you they are.
Where I work, and in my college, people are more disciplined than most who come (came, before the 5$/month filter) on this forum. And oftentimes, they are more knowledgeable too. I personally don't understand why in self-improvement, where I've discovered about entrepreneuship, being an employee was treated derogatory. I've stayed away from employment for years and pushed on business without becoming an expert first and it hurt me bad. I wish I didn't.
#6: getting and working an expert job for extended time will teach you much more than years of being a failing solopreneur
Fact; it's hard to get a job, and it's even more complex to get paid consistently later. You need to prepare months ahead, and you need to work 40-60 hours a week to stay. Maybe more. Because somebody else will take your place. And you will panic because if you don't deliver by a deadline, you won't be paid.
But you know what? You may not be paid a lot in the beginning, but you are being paid to learn from your own mistakes too. No amount of trial and error in a business has told me as much as a job did.
#7: Switching projects is deadly. And there is pressure to switch.
Building projects is like building a rocket - it may fly, but it certainly won't if you run out of steam before you can complete it.
Common lists of problems pushing you to switch a project:
- You are out of money to continue, so you postpone it and never come back,
- You don't know how to ship X,
- You are getting issues from elsewhere (personal or otherwise),
- You realize that the project shouldn't have been started in the first place.
And you quit midway. But the rocket that you've built will not fly. It will rot, and soon enough, you'll need to build a new one.
"rockets" may be skills and domain knowledge too.
#8: Gains from "outworking others" are not real if you are switching projects:
This. People who are on the top of their fields - that I look up to - have compounded their work - and I mean real, verifiable work like:
- publishing papers if you want to do a PhD,
- having income,
- having company names on your resume if you want to have a job,
- going out there and trying things if you want to start a company,
- selling to others if you want a sales position.
If you switch 10 projects in 3 years like I did AND you aren't an expert in your field, you will achieve nothing but wasting months of your time (and in case of software, stressing servers of Github).
#9: When you start a business, you will stop learning "hard skills".
Tying into #1, #2, #3, #4:
You won't or you will spend incredible quantity of time learning hard skills and domain knowledge before you can turn a profit. If you want to start a software company (picking an example from my day job) doing analytics with AI for ecommerce businesses, and you didn't have experience in both software AND analytics AND ecommerice, you won't complete it. Where I work, for example, the only reason we do have it is that:
1. Our founder had already owned a ecom business,
2. He hired great talent for his work, and I'm talking of our technical leadership who work 80 hours a week for a not very high wage; maybe equity.
What's worse, is that if you try to learn either or both of the two, by the time you will in fact finish learning, the marketplace will change and you'll need to continue learning more; and when it'll be time to serve your product, your knowledge will still be shallow.
Citing #4 - you won't make money if you are not a domain expert. In general.
#10: You can't fallback to a well-paid job that you want to quit to if you fail your business.
I heard advice that you'll have higher salary if you start a startup. From multiple places - YCombinator, and others. This is a lie. In fact, you'll be treated as having a career break, because it is in the eyes of the employers. So that job at Google that you want to take after starting a company? no no.
#11: STEM degrees pay (unlike what MJ says.).
I don't know one would advise against colleges - probably because in the US they are expensive. It appears to me that people who advise against degrees didn't pick the right ones.
I never had somebody advise me against a degree in-person, though sure I've heard lots from the internet (including from books many of us have read)
yes, it takes learning, but you are far better of having a proper salary right after college and subsequent years - and in STEM the salaries are good. Having a good salary allows you to do business, and being an expert that others want to work with. If you are a plumber, you have the same pay as somebody from college, but don't have the privilege of #2. And you don't have the expertise necessary for starting anything but a plumbing business because #9.
#12: You can't start a company if you don't have capital.
8/12 of my business attempts over the last 4 years stopped because I couldn't continue financially. Get a paid skill before you start a system.
#13: Wealth is built incrementally, and I've never seen anybody do the first few steps from a tech business.
If you are there like me building a company after college, you will not be able to hire others. And that's preventing you from getting skills you don't have or buying technology that you need. For example, I couldn't buy enough AI compute to finetune AI models properly, because those needs hundreds of dollars. And I don't have hundreds of dollars lying around. Moreover, I don't have others to hire to streamline the process.
citing MJ's video, I subscribe under every his word "From Flat-Broke to Filthy Rich: How to Get Started Toward Wealth When You Have No Money & No Skills" :
That's it from the learning that I had.
The rest of what I learned were only technical skills, really. Nothing that most people will care of on this forum, as far as I understand.
How I spent the last 4 years; and how I shouldn't have, had I known these:
Now, to back things up, below is my story. Had I the thirteen points above, I wouldn't go down the road of spending 4 years to learn all of this.2021 (not interesting, but useful for context)
In 2021, I have just finished school with bad grades (back then, I didn't understand why somebody would like to learn... especially since everyone in my family had Master's and my parents both had an income below $1000... which was low even for my country standards. For no reason in particular, I disregarded grade school. I'm from Ukraine, and by having connections with officials and knowing that war was coming, my family commanded me out of the country to a degree in Europe. My family didn't want to sponsor a degree in a good college or country (reasonably so.) so I went to a below-average college in Slovakia which IMO (and to an opinion of most studying there) was more of a visa mill than a college.But I did try in that college. By the end of 2021/2022 I was the top of the class because I was sick of bad grades and for the first time, decided to learn.
2022
I've learned about the entrepreneurial path in the summer of 2022 when I was 17; courtesy of a self-improvement YouTuber. Just before I learned about "the Fastlane", I have decided to learn software engineering because I started a business degree and they said salaries were 3x in software - and innately, I wanted to be an engineer and work of business managers/HRs just didn't resonate with me.Just 5 months later after I started learning software engineering, I got a job paying 8$/hour - well above average for Slovakia, where the average was 4$/hr.
But I felt pull towards business and disregarded that job. I planned out my business when I was supposed to be working. moreover, I didn't even understand what the product we were developing was about. I had a 2-hour zoom meeting with an engineer, but I was too green to understand that that was enough to start working; because I didn't know how to work.. I was able to write and read code, and did ship two or three PRs over the course of the three months I was on the job, but that was it - i didn't even understand what problem the project was solving. I understood it just a few days before I was fired.
I did code my software project meanwhile. I actively spoke to users (every day, found conversations through Reddit). I wrote the software - it was difficult and I had to learn multiple languages, but I did. and I learned a slight bit of web design. The truth is, I wasn't an expert not at the software not at the domain I was building it in; I had no idea what "marketing" meant or how to even sell something to somebody - I never had any experience selling, or working anything.
So I just developed it day in and day out. The result was so-so because - I wasn't an expert.
I've listened to YCombinator videos and listened the promises of them telling that you can become a billionare in the field and that it's better than a job. Yeah. Except sometime until 2026, the job I was fired from was the last income I've made from software for 5 years.
I don't have much to inject from what is already said, but this sets the starting point.
(retrospectives #1-4, #7, #9, apply)
2023 (2023 and 2024 were laughable? I didn't know what I was doing. I'll keep it brief.)
Starting companies. FullCook - a machine that can build anything you want with a press of a button. Notably, I had no engineering experience. It took a to even design something remotely viable: MARKETING - FullCook: how to go about pricing?I was on family money support.
(retrospectives #1-4, #9, #12)
2024
Made a patent for FullCook.Decided to stop a long-running project of FullCook (on which I worked for 10 months). Decided to do an AI that will change the world (quite literally). The idea: make an AI that can be a hardware engineerz. It didn't exist, I've double-checked. I spent 4 months doing an environment in which the model could be trained. My Bachelors degree is over.
Moved to Ireland. In Ireland, lived for welfare because we were given it. Shame, but I spent a year on a welfare doing this project. I think that had I not stopped on FullCook, I could have continued it as a project and it could be a meaningful business by the end of 2024.
(retrospectives #1-4, #7-9, #12)
2025
First half of 2025:
I've Learned AI and Python; software engineering as a whole. For real this time, from the projects and many models that I trained while learning.
After learning machine learning, software engineering, switch projects. Started more startups, spending a month on each. A project in robotic assembly which I never finished because again - I had no money and my knowledge of AI was still shallow. A project that was meant to accelerate training of AI/LLM models. Also didn't finish or publish it because I had no money.
Second half of 2025:
Welfare money ran out mid-2025. I moved to Dublin: I was certain that I would find a job and there are more money in Dublin, even in hospitality, but I couldn't. - AI - - AI entrepoorneur's resume for hospitality jobs
Money has almost run out. Started window cleaning again! Doing 200$/day - WOW!!! most money I've made in my life. Paying my own bills.
My family pushed me get a degree, which I enroll into. Tentatively, I get a Master's degree for 9000$ in tuition paid in installments over the next year. It started in January 2026 and lasted one year. It appears as a OK college.
Continue starting 3 more startups in all my free time.
I started daydreaming of becoming a software engineer at Google or in any appropriate company - after all software engineer are making lots of money - in Google they said 140-300k/year depending on seniority + equity.
I'm hoping for the best, and I'm more software as startups until November
October 2025 and I'm cleaning windows.
At november and December, I've committed to window cleaning full-time. Wake up at 8am, start window cleaning by 10am, finish at 8pm. Watch youtube videos which talk about it. Made something like 5k/mo working 6 days a week, hired an employee, ran ads, bought new gear - lost half of it.
(Retrospectives #5, #9 and #13)
2026
January:I started 2026 by paying off my college tuition. Saving money from window cleaning, I've paid off everything that I had for two months for college. I resumed did window cleaning from January 5th to 15th, but suddently, I was making 2x below minimum wage - 5-6 EUR an hour. Nobody wanted their windows cleaned in January! For days straight, I came but with nothing in my pockets, soaked in rain. By February, I was out of savings.
I've paid 120$ for a Jobber CRM in November, and I hated every bit of it. it took me 4 minutes to enter details of a client! so having free time and not doing window cleaning, I decided to make a project doing a CRM which would be much faster to use than jobber was. The idea: text all messages into Whatsapp instead of a CRM. You could type something like "add client john window cleaning, 50$, Millstreet 5" and it would log a job just as a CRM would. I made a UI too.
I've stopped after two weeks of intense full-time developement because we needed to do a project in school and I decided to revive my idea from 2024. I also wasn't sure anybody would actually buy the product. Nobody ever knew about the project until you reading this.
February:
We had classes of AI. Because I've spent the whole 2024/2025 learning machine learning I was fairly competent and well, I didn't learn much. I always made it a point to show up and comment and be active in the class.
We had a project due by March. So I decided to revive the engineering AI startup, trying to make a publishable research (a paper) about it, on how to evaluate and how to improve capabilities of LLMs on it. It took me whole 2 months of in and out of developing it, and it barely worked at the end. Actually, it did work as intended, but LLMs were in fact incapable of solving the problems - not a single problem was solved within a time budget. Mind you, I had no money to invest any further, so the project was shutdown.
I've also gotten an offer from an engineer to join a startup, where CEO offered free hours and remote job. I hesitated - will I come back to a job after success in window cleaning? I've spoken with a CEO and the contract was a unpaid internship.
At that point, I've written code for two years with 0 income from it, so I accepted it.
I've reengineered their entire application - their backend, UI, added a pile of useful features, and left it for their review. Incrementally allowed them to code review it. they responded after 3 days after each of my message. At some point, I realized the startup was in fact running for two years and had engineers shipping features very occasionally.
I was invited to be paid for my work - 1.5k$ for what for me was 10 days of work. I said that - maybe I should become their technical co-founder? they didn't say yes for no reason in particular. So it was a no.
Restarted window cleaning on March 15th. Consistently running out of savings though because college + own startup + window cleaning + working for the startup - the window cleaning didn't fit in. Needed money support for tuition for my parents. Always paid my own rent though.
March:
Money-wise, march was the same. Window cleaning pays bills, pays tuition, pays food. I am developing the AI paper.
Two people startupschool.org have contacted me during March. Two potential co-founders. One suggested doing a software project together and one asked if I was interested to contribute to his project.
The former was a business owner with a track record of selling websites and marketing to business owners. Said he and his co-founder were scaling up a marketing agency to up to 30k profit/mo in one month. then they part ways with his cofounder, because his cofounder wasn't interested in software. He looked on me to start a software company with him. I was interested, so I did accept it. We discussed it. I went to different city to speak to potential clients. The clients informed us that the idea was executed many times already; in particular, his idea was software for quotation of houses. So I have reinforced that because there are literally >20 companies doing the same, we shouldn't do it too.
April:
I've gotten an offer from the "contribute to his project" startup founder whom I've met at startupschool.org. He turned out to be Russian. He offered me to join his company (Nyle AI Unmatched Tool for Amazon, Shopify and TikTok Shop Sellers) for 2k$/mo. He said: "We need to ship features very intensely for 2-3 months and then I'll be able to raise wages because features will be built". I've hesitated, asked for advice from software engineering roommates who did 90k$/yr (fine by Irish standards) and was advised to take the job. I did take the job. This was my first real job.
I passed an interview due to my machine learning skills; where I did the system they asked for in a day.
The founder has promised strong salary gains if I performed well.
The codebase had no documentation whatsoever and I had a contact with only the lead engineer and a non-technical founder.
I immediately jumped into vibecoding the features they wanted or needed, talking directly to the non-technical founder. Merged 10 PRs after 14 days. Felt good about myself.
Retrospective #1: wow: real jobs are real.
Retrospective #2: Wow, did I get paid money to write software for the first time in two years?
Retrospective #3: I still wasn't explained what the project was about properly, and was outside of the team for 14 days,
----
On starting own software company side, I paused it or contributed to it for 10 hours a week because I had a job and if I wanted to be somebody like that 90k/year roommate (and I wanted to), I knew I needed to push my job hard).
May:
It turned out, my new code broke a few things and a lead engineer had to fix some of it. It turned out, we have had >1m LoC in the codebase. There was real history of development. Unfortunately, it was not at all documented. It was just a pile of code and a repository from my point of view. which it was - kept in heads of two or three backend engineers.
I got integrated into the team. Turns out they had a proper workflow and proper everything, but the founder was not technical enough to know or care, much less to tell about it. the lead engineer was mute too. I couldn't have known that the issues I caused could be issues at all, but a particular external API we were working with had underlying issues too, and I have unknowingly tripped them. Had to undo much of my first months' development efforts.
after the first few weeks of being integrated with other engineers, I have for the first time shipped something useful to somebody. for real. Took rework, but I did.
Now was the time to be proud of myself.
And get paid.
I started noticing a pattern, though:
- I was working 7 days a week, 6am to 9pm.
- So did our lead engineer,
- Everyone on our team is online and working 60-80 hours a week.
- For 2-4k$ a month as I learned later.
- Also, the company founder we were with has pressured us hard at all times on shipping even faster. His derogated everyone's work in front of everyone on every daily meeting. What's happening? it was fine during 1-2 months - I closed eyes, but this was consistent.
---
starting a company side:
We paused for two weeks after I explained to the cofounder that the idea was in fact dead before starting. After two weeks, he reached out and pitched something else - to start a company building websites powered by software. Basically, that's what I did. He asked to collab with me on the repo, and I granted access (he was a better designer than me; I'm a systems engineer, not a designer) and after 3-4 days of prototyping, I shipped the first bit of software for him. It looked not the best, but it was sufficiently functional.
Then he turned off his messenger for 4 days,
afterwards, I start calling him up through any messenger I can... and get an answer:
He took my code and ran away.
What?
June:
Job-wise, June went by the same - working 7 days a week all available hours.
I was told to start a project which sounded fairly simple and I executed the first iteration of in a week or so. Then requirements changed. I needed to rework the whole thing. It took me whole june and more to rework it.
---
I stopped window cleaning. While the cofounder has ran away, I start making on another project, a window cleaning machine which I've detailed a while ago on the forum:
I've built up a simulator environment (much higher fidelity than this mockup too, where I have only experimented with pole movements) and while I wouldn't upload the whole video here, the robot could drive in a realistic environment and clean windows.
At some point the co-founder came back. I said that I'm willing to continue - just let me re-engineer a few things because I was in a rush and the first codebase was sloppy.
We agree on the following: I keep the codebase outside of his access.
I dropped the window cleaning project and was booked up again. with job + startup.
---
July:
Job: I'm stuck reworking the thing for a whole month because the requirements have changed. I didn't deliver anything meaningful for a month. I'm starting to panic. I worked for 7 days a week and got paid 2000$/mo. This is minimum wage in Ireland. I'm out of savings because I need to pay tuition too. I needed to restart cleaning windows on Saturdays.
I devoted the whole July to the job.
August:
Job: the whole project which took 1 week to do by old requirements now sprawled into 80K LoC because they asked for new features. It became unworkable. At august 1st I needed to delete all code because new features that were asked in June needed a completely different approach that rendered the first unviable.
At august 10th I'm asking "so where's my minimum wage which should've arrived at 1st, boss?". To which I'm replied "where's user features? where can I see your work?"
And I couldn't answer. I needed to rework the whole project because they hadn't told me all requirements from the start.
At august 20th I'm still not paid for July.
I had a planned vacation that was booked for my family for two weeks (August 24 to September 6th), so we flew off to the ocean. I couldn't finish reworking the major feature by that time, so I said I'm going on the vacation and will finish it afterwards.
Startup:
Since i'm on a vacation, I devoted 2 weeks to the startup. Literally polishing every small bit so that no part of the application will be slop, which would foregone the need of rework. Polishing every small part before the implementation. See: the last project was stalling because the codebase has become such a nightmare to work with due to rushed implementation that it become outright unworkable. I couldn't comprehend - or could, but with real problems - on what was happening in the codebase. It wasn't planned out, it was LLMs doing decisions and then getting tripped up by them and reworking everything.
This time I decided to do it differently. I spent 2 weeks writing 90 markdown files total outlining every small bit per feature so that nothing, nothing can go wrong when I implement.
My vacation lasted for two weeks; which I spend specifying how it will be built. By the way, this pays: you can reverse decisions in markdown 10x faster than you can in code. Better yet, you'll actually comprehend them. You can ask a model along the lines of "are there any issues" and it'll come back with an answer in 2 minutes, not in 5 or 7. You can make edits 10x faster. but your app isn't built.
September:
On Friday, I'm getting this message:
I reply that "I didn't even start coding yet". The application is very well specified; I removed all technical debt before starting. Minor bugs remain, which I proceed to patch in doc-only mode. By Sunday, I realize while the rest of the application (authentication, user management, other functionality) is very well-specified, the website editing itself needs some rework.
at monday, I'm getting a message:
What? I've just spent two weeks and we are days away from launching the product - since the application is implementation ready from a spec; I'm confident there is very little to be left to do.
I get on a call, and he says that I must wait until we meet. And that we must meet on Friday.
Two days later:
Well, I wait. I am only a technical cofounder in this business and effectively have no control whether the business is moving or not moving.
---
Job: I still wasn't paid. Will maybe be paid when I finish the project.
Our founder has held salaries for me and another engineer. I speak to another engineer, and he says he works 250-300 hours a month per his estimate for $3000. He was also, told that he would get a raise based on performance. which he never did. In fact, he had offers from better places, but our founder convinced him that "a few months and we'll get rich, and you will have high salary".
I need to continue doing window cleaning to pay rent (I'm already past due because I was on a vacation for 2 weeks). I do.
At Friday (today), I'm getting a message from the co-founder that we move over to Monday, so we postpone the meeting.
So what's happening?
---
Meanwhile, my family pushes me to get a PhD for two reasons; safety and brains. I think this is worthwhile. Given that currently I'm priced at 2k$/mo as a software engineer - and no other job offers, this appears like a sane thing to do. After all, I wanted to work in places like Google Deepmind or OpenAI or Anthropic.
I hope the story will serve useful for others - my goal was to have others to not make the mistakes that I did. For those above me, if you will ever have influence over others who want to do the mistakes that I did, please refrain them from this, for these are mistakes that cost real % of human lives.
Open decisions:
1. The most pressing one: I have taxes to pay by mid October. about 4k$. I can either try to work my job and finish a project which I will be paid 4-6k$ for - I asked, but our my employed eluded a clear "yes" for 3 times that I asked. Or do I do windows? probably a combination of both. Windows take a full working month to earn this much - when I earn 250$/day. And the reason why I need it, is because I'm a foreigner in Ireland, and this year they said whoever doesn't declare 30k$ in income, may get their visa not extended. It's not a hard fact, but it may happen. And i'm not going back to Ukraine, that's a surefire way to get multiple bullets into your body. (Many of those who I knew are killed already.)2. Should I switch to being a software freelancer instead of looking for jobs? I've heard some have success this way.
3. This is a wrong audience to ask, but is PhD a right choice? for AI research, presumably yes. It may also open doors to any country in the world. Best of all, if I'm a student, I'm not getting departed if they cancel all our visas, and that's worth more than anything - that's an existential question at this point (that's why my family is pushing me anyway).
- If I want to apply to a PhD I need to publish research. And if I need to publish research I need to work on research. And that takes away time from job, freelancing or windows.
4. Should I stop doing any business for the next 5 years and commit to increasing my income through other means any means necessary? this appears like the right choice, given my own findings. then, PhD is conflicting with it.
5. I've already invested a month of work from the aforementioned co-founder. Do I continue? Sunken cost fallacy, but not having control over that what I need to spend weeks on is damn scary. We have shared understanding. But he says "stop", it stops. It happened twice already. but another dead project?
6. I also need to start saving up for a house in Ireland, which is well above 200k even in poorer regions. Forget about CENTS until I have a house and good passport?
I don't think I have the capacity to pick anything but a) finding a better job, b) doing a PhD, c) doing freelancing (either of, given how time-consuming all 4 are) d) doing window cleaning further to pay for tax. Startups appear out because I'm booked up until the tax is due. And maybe I should stop trying in the first place. Also, I need to prepare for the next January and February, for I will have no income if I won't have savings.
Dislike ads? Become a Fastlane member:
Subscribe today and surround yourself with winners and millionaire mentors, not those broke friends who only want to drink beer and play video games. :-)
Attachments
Last edited:
Access Restricted: Unlock 1,000,000+ Posts
Stop Peeking Through the Keyhole.
Open the Door.
You hit a wall because the best advice isn't free—it's earned. Since 2007, MJ DeMarco has been active here daily (99.9% active rate), building a war room for entrepreneurs who refuse to settle for mediocrity.
- Active Daily: Life-changing content posted dozens of times every day. No dead threads.
- MJ's Inner Circle: Direct access to the author of The Millionaire Fastlane.
- Vetted Network: Connect with founders scaling to 7 and 8 figures.
- Proven Roadmaps: Strategy over motivation. Execution over "hustle porn."
"SCALED TO 6-FIGURE MONTHS"
"Tips and advice here saved me from huge mistakes, others allowed me to scale my business to 6-figure months."— User MitchC
"INSANE QUALITY OF PEOPLE"
"The number of high quality people I've met from this forum has been insane. All of it predicated on building real value."— User Richard Peck
"You are the average of the five people you surround yourself with."
Are you surrounding yourself with success?
