Stan Tatarnykov

It'll post a nice list of issues so you don't have to think about QA. You just come in and fix what's broken. And you keep a dock where it'll take notes from the last iteration. You can run this once a day after all your changes. Saves me a lot of time.

People who are ignoring AI are like people who were ignoring the internet

at a job, someone picks your work. boss, PM, team lead 99% of decisions are made before you sit down solo, every decision is yours what to build, when, how long, whether to scrap AI makes it 10x worse — every prompt opens 3 new paths you're not lazy. you're drowning in options

every founder i know feels guilty about vacation but when you're grinding every day you stop asking the most important question. am i even building the right thing? grinding keeps you working IN the business. shipping features, answering emails, putting out fires. no time to zoom out every time i step away i come back with a clearer picture of where the business should go the guilt shouldn't be about vacation. it should be about spending months executing without ever checking if you're pointed the right direction step back. look at the big picture. that's not laziness, i see it as strategy.

Most people still have absolutely no idea how to use AI. If you spend a day on Twitter, it makes you feel like you’re behind but you’re not. At all. chatting with a few non-tech friends made me realize most regular people barely even tried using ChatGPT, let alone something like Claude code. There’s a huge disconnect. We’re still super early in AI. I posted this video on more mainstream platforms to get through to people.

this one mistake cost me 2 years and hundreds of thousands of dollars in my startup after my game hit 80M players I let my ego take over. hired 4 people, kept the project secret, built for a year and a half could have built a simple version in a month and learned the same thing two signs you're doing this right now: 1. you're keeping your project secret because you think people will copy it (they won't, nobody copies unproven ideas) 2. when someone asks who it's for you say "whoever likes it will like it" please learn from my mistakes

Vibe coders are just slot machine addicts. Hear me out. Every prompt is putting money in the machine, hoping to hit a million-dollar startup. Meanwhile Anthropic, OpenAI, and Cursor are the casino... they get paid no matter what. I went to AI events. Talked to tons of builders. Almost nobody had talked to a single customer. Because talking to customers doesn't give you dopamine. But building features nobody asked for? That does. Skip the build. Validate fast, kill bad ideas faster, and only build what people will pay for. (I sold my game for 7 figures in 2020. Now going all in on AI. More coming.)

I built a game that hit 80M players. Sold it in 2020. But AI took over game dev now... So I built an app where you scroll other people's mini apps and make your own in 30s. minis.im Someone made a game where you play flappy bird by doing real pushups. Just launched on iOS. Would love some feedback.