Arjun Jain

Costco food court is undefeated

Anyone can build anything. A real engineer has opinions and design taste.

All kunal and I care about is winning

Two types of people (both valid): 1. Work is their passion 2. Work to do their passions

Sell sell sell

There is this mystical fog hiding amazing talent. Idk why it’s there but it’s there

@VeerCumulus You know me too well

@Arjunjain1017 bro just tryna flex having friends

Love it when my friends at open ai and deepmind argue over which models are better. I have no friends at Anthropic

Should I switch from superhuman to avec?

Hardware is the other half of software’s story. As software races ahead, the cracks in our aging hardware design tools will only become more obvious.

Just interviewed someone using cluely. It’s shit

Friend or foe

Tonal dis you?

@gabednconfused Can you share the 1 page reference guide with me pls

We know how to build robots that lay bricks, finish drywall, and print concrete. We have not figured out how to build the companies behind them. The platonic ideal for investors funding these companies is to see you start as a vertically integrated subcontractor, prove the unit economics, then shed risk over time until you become an equipment manufacturer. It is a clean arc. And as a destination, it is probably right. Caterpillar sells machines. John Deere sells machines. That end state has comps and is defensible. But the path to that destination is never clean. It is a river. And the river is shaped by two things most founders get wrong, identity and capital. Most robotics companies have a natural groove. It is set early and it is hard to fight. Some companies are OEM-leaning. The founders invented new categories (3D printing, autonomous finishing, heavy equipment EV) and the hardware itself is the breakthrough. Manufacturing and engineering is the core competency. Other companies are services-leaning. They use the best available off-the-shelf robotics to automate a trade. The differentiation lives in the efficiency of the operation and a consistently lower cost output. This identity should set the direction of your business model. An OEM-leaning company that forces itself into services will feel friction at every turn. A services-leaning company that tries to invent breakthrough equipment instead of solving field problems will also struggle. Fighting your groove early costs years and capital that most startups do not have. The companies that fail usually have working technology and willing customers. But they die because they cannot attract enough capital. There is effectively one pool of risk capital available. Venture. You are asking VCs to underwrite unproven technology in a conservative, slow-to-adopt industry with capital-intensive assets. That is a lot of faith per dollar. They can only stomach it if each dollar you spend buys dense, legible proof that the next dollar will be easier to raise. But capital does not move on data alone. It moves on narrative. Every fundraise is an escalation. You are telling investors that the thing you proved last round implies something larger this round. If the narrative is right, each escalation feels inevitable. If it is wrong, each escalation feels like spin. Evidence without story is a science project. Story without evidence is fiction. Investors need to see where it is all going and need to know the next buyer will be there. There will be winners in construction robotics. The market is too large, the labor shortage too structural, the equipment too stagnant for there not to be. But the winners will not be the companies that followed the platonic ideal. The density of proof you need at each stage of financing is always bigger than you think. The winners will be the ones comfortable doing more than investors expect, making each round's proof impossible to ignore. "If you do everything, you will win." — LBJ

We're hiring founding engineers at Tandem. Come build the system that helps hardware teams make better engineering decisions as products get more complex. San Francisco | $120K–$250K + equity tandemai.io/careers #hiring #ai #tech #startup #sf #ml #llm #foundingengineer

[1] for the last two years, the conversation has been about individual productivity. give everyone Copilot. ship faster. write more code. but the real shift is much bigger than that. it's not about making people faster. it's about redesigning the entire org around agentic systems — and the companies doing this are starting to look radically different. 🧵

You can now enable Claude to use your computer to complete tasks. It opens your apps, navigates your browser, fills in spreadsheets—anything you'd do sitting at your desk. Research preview in Claude Cowork and Claude Code, macOS only.

Figure 03. Impressive. 🤖🦾🦿@adcock_brett

@ritwikpavan @shivanijpatel @DShankar @mehul @nathaliegou @ragsgups @thatssodhawal @EvenRealities Would love to join!

We’re renting out a bar this Friday (3/20) and bringing together 30+ hardware founders, including: - @DShankar - @mehul - @nathaliegou - @ragsgups - @thatssodhawal Comment below if you’d like to join!