Mark

@daniel_dhawan how did you find your first paying customers?

At my first startup, I had 200K users and almost $0 revenue. At my current startup, we scaled to $1M ARR in 3 months. Here's what changed: First startup: - 150K users, almost zero revenue in 3 years - I thought users = success - Did everything myself, moved slowly - Obsessed with the product. Ignored the business At Rork: - Built for a problem I personally struggled with for years - Found the best co-founder @Levan to move faster - Launched before it was perfect. Charged from day one - $1M ARR in 3 months after launch Most founders I know don't fail because they build something bad. They delay monetization, iterate slowly, and by the time they figure it out, the runway is gone. Don't repeat my mistakes - charge early and ship fast to succeed. That's it.

@itsAlishaPandey all in on 1DTE nvidia options

Just bought a new ThinkPad. Now how do I make $1M/month?? 🤔

@Vatsalpandya333 lets gooooo

7 days until I drive 1,600 miles to San Francisco. You don’t realize how much of your life lives in a place until you start packing it up. The chair where I took every Zoom call with my parents. The desk where I made my first design partner pitch. The Walmart whiteboard that turned into 8 months of architecture. Mumbai to Nebraska at 18 was about survival. This move is different. SF isn’t a city. It’s a test. A test of whether the bet on myself was a real bet or just a story I told to feel brave. 7 days. 1,600 miles. 4 founders. 1 house. 0 backup plans. Let’s go.

@tom_jiahao Cost, high resolution tactile sensing

What matters most when buying a high-DoF robotic hand? Payload, sensing, control, reliability, price, sim, repairability? We’re working on something big. Your input may shape it.

@tom_jiahao you only need 3 fingers

These "complex" diffusion theory stuff is super simple/intuitive once you have the right mental model (like 3blue1brown type thing). There needs to be a benchmark or measurement of how good agents are at educating humans on complex topics. I think this is super important on alignment and productivity of human-agent systems too

@dreamworks2050 @TheAhmadOsman how much does power cost for this lol?

@TheAhmadOsman OH FUCK YEAH MY BODY IS READY!! 🔥

DOMAIN SECURED. WE GO LIVE THIS FRIDAY. The LLM GPU Build Guide > 1x to 8x GPUs > $2K to $15K+ budgets > inference, training, cpu offload > full parts lists, tradeoffs, where to buy & everything learned the hard way > PCIe lanes & bandwidth, why 2^n # of GPUs matter & more

@Star_Knight12 ngl babysitting agents is really hard work

2022 - Student 2023 - Developer 2024 - Prompt Engineer 2025 - Vibe Coder 2026 - AI Agent Babysitter 2027 - Farmer

@CodeWithAmann understand the code

Software engineers, what’s your plan B if AI writes better code than you in the next 3 years?

@metrox_eth did you make the hardware yourself?

First dataset, first training, first eval, all working. Jarvis is rough but this is the floor, only goes up from here.

@TheAhmadOsman @micheltamanda Is there a tutorial or blog post of how to get started? (ideally on a budget)

@micheltamanda Just for the curious, here is what started my Buy a GPU, The Movement x.com/TheAhmadOsman/…

Buy GPU's said @TheAhmadOsman, so I did and had no clue what I was doing. 3 months later I have a fully local AI research operation, a solid understanding of AI, and confidence that AI will not take my job. I ask you, for the sake of humanity: buy a GPU, learn to host your own intelligence. Thanks to those who I learned from: @steipete , @0xSero , @Teknium , @LottoLabs , @NousResearch , @TheAhmadOsman , and many more I forgot to mention.

@TheAhmadOsman Bro is a landlord while Im stuck paying rent

My name is Ahmad and I have a Compute problem

1. One can not outsource understanding. 2. Understanding is important in longer horizon problems. 3. Therefore a good way to vibe-theory is to generate small verifiable chunks with AI and fully understand it (similar to speculative decoding).

@Palak3312 starting with a microcontroller is more beginner friendly than an fpga xD

RIP my robot 😭 LDO didn’t survive… redesigning the PCB now

@Tech_girlll this is so true. found out the hard way

Why is no one talking about this?

@AntoineRSX I think we should build instead of six packing

For a long time, a six pack, athletic body functioned like a social signal of self-command. It suggested routine, restraint, patience, and goal-orientation. Now physique becomes increasingly accessible through peptides, surgery, and optimization tools. It's losing its status symbol. I wonder what the next “six-pack” will be.

@ryanndngg but what do you actually do tho

We're racing to make Skalpel AI the fastest company in history to $1M ARR. I need engineers. San Francisco. This week. We're going to break the record. Or die trying. May 15th launch. The rules: -in-person interview only. -no zoom theater. -you walk into our office. -we pair on something real for 2 hours. -same-day decision. DM me. i'll send the address. we move tomorrow.

@justinnguyen damn I have my own wiki but didn't think of the generational thing

Hi, since a few people have asked. “Knowledge preservation continuity infrastructure” sounds like word salad I know. What I am creating and helping users to build is something called a “Stratum” which is a collection of your own personal notes in markdown format just like in Obsidian. In Mithrify, the notes will be will be indexed and placed into categories or domains that make up who you are. Things like your identity, memories, skills, knowledge, interests. Then takes it even further by reading what patterns develop over time such as your decisions, your cognitive thought processes, judgments and behaviors. The more notes you create about yourself, the more detailed your Stratum becomes. It essentially begins to become your own personal knowledge base “second brain” as others have used. If you read Andrej Kaparthy’s notes on the LLM wiki, it’s basically like that. I’m going further with that and creating a way for users to transfer their knowledge down to their descendants (your kids, grand kids, so on and so forth) so that they can look back at any of their ancestors Stratum’s to access knowledge and skills that descendants may never have had a chance to learn from directly. For eg. Let’s say your mother knew how to create a specific dish. She cooks it a certain way, with certain ingredients. She can transfer her knowledge of how she cooked it into a markdown file and have it available to any of her descendants down the line since they can look back at her Stratum. This preserves knowledge across many generations.

Submitted my application to YC. No pain, no gain. No guts, no glory.

@nico_laqua The adversarial selection is going to go crazy

Today, we're excited to launch AI Coverage (insurance for when your AI messes up). Insurance was built for risks that have existed for decades. AI is creating a new category very quickly.

@ndrewpignanelli At least for research / technical things, you cant not look at the code for the important stuff. Learned the hard way :( Probably fine for non-technical businesses tho!

Announcing Cofounder 2: Run an entire company with agents. It's the infrastructure for the one person billion dollar company - orchestrating agents across engineering, sales, marketing, ops, and design. (and yes that's my real grandma in the video)